CN119904166A - Information prediction method, device, equipment and storage medium - Google Patents
Information prediction method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN119904166A CN119904166A CN202311405589.9A CN202311405589A CN119904166A CN 119904166 A CN119904166 A CN 119904166A CN 202311405589 A CN202311405589 A CN 202311405589A CN 119904166 A CN119904166 A CN 119904166A
- Authority
- CN
- China
- Prior art keywords
- target
- item
- items
- returned
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
- G06Q10/0837—Return transactions
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Data Mining & Analysis (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Artificial Intelligence (AREA)
- Development Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Resources & Organizations (AREA)
- Evolutionary Biology (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
本发明实施例公开了一种信息预测方法、装置、设备和存储介质。该方法包括：获取仓库中的退货物品；获取当前迭代周期对应的目标物品属性和目标分类网络模型，其中，目标物品属性是基于当前迭代周期中的样本物品对应的待选物品属性信息和实际破损等级，从待选物品属性中确定的物品属性；目标分类网络模型是基于样本物品对应的目标物品属性信息和实际破损等级进行训练获得的；基于目标分类网络模型和退货物品对应的目标物品属性信息，对退货物品进行破损程度的分类预测，确定退货物品对应的目标破损等级。通过本发明实施例的技术方案，可以实现退货物品破损等级的自动预测和准确预测，无需人工参与，提高了破损等级预测的效率和准确性。
The embodiment of the present invention discloses an information prediction method, device, equipment and storage medium. The method includes: obtaining returned items in the warehouse; obtaining the target item attributes and target classification network model corresponding to the current iteration cycle, wherein the target item attributes are determined from the attributes of the selected items based on the attribute information of the selected items and the actual damage level corresponding to the sample items in the current iteration cycle; the target classification network model is obtained by training based on the target item attribute information and the actual damage level corresponding to the sample items; based on the target classification network model and the target item attribute information corresponding to the returned items, the returned items are classified and predicted for the degree of damage, and the target damage level corresponding to the returned items is determined. Through the technical solution of the embodiment of the present invention, the automatic prediction and accurate prediction of the damage level of the returned items can be achieved without manual participation, thereby improving the efficiency and accuracy of the damage level prediction.
Description
Technical Field
Embodiments of the present invention relate to computer technologies, and in particular, to an information prediction method, an apparatus, a device, and a storage medium.
Background
With the rapid development of the e-commerce industry, more and more users like to purchase articles on an e-commerce platform, and the users can return the articles if the users are not satisfied with the purchased articles, so that the purchase experience of the users is improved. The reverse logistics is more used for the e-commerce platform to process the returned articles of the user.
Currently, an e-commerce platform generally performs manual judgment on damage conditions of returned goods, and performs subsequent processing based on a manual judgment result. The e-commerce platform can also directly return the returned goods to the goods supplier for processing by the goods supplier.
However, in the process of implementing the present invention, the inventors found that at least the following problems exist in the prior art:
The existing method for judging the damage condition of the returned goods by means of manual experience is time-consuming and labor-consuming, and the accuracy of damage judgment cannot be effectively guaranteed.
Disclosure of Invention
The embodiment of the invention provides an information prediction method, an information prediction device, information prediction equipment and a storage medium, so that automatic prediction and accurate prediction of damage level of returned goods are realized, manual participation is not needed, and the efficiency and accuracy of damage level prediction are improved.
In a first aspect, an embodiment of the present invention provides an information prediction method, including:
Acquiring return goods in a warehouse;
Acquiring a target article attribute and a target classification network model corresponding to a current iteration period, wherein the target article attribute is an article attribute determined from article attributes to be selected based on article attribute information to be selected and an actual breakage level corresponding to a sample article in the current iteration period;
And based on the target classification network model and the target item attribute information corresponding to the returned items, classifying and predicting the damage degree of the returned items to obtain target damage levels corresponding to the returned items.
In a second aspect, an embodiment of the present invention further provides an information prediction apparatus, including:
The goods returned acquisition module is used for acquiring goods returned from the warehouse;
The system comprises a model acquisition module, a target classification network model, a model analysis module and a model analysis module, wherein the model acquisition module is used for acquiring a target article attribute and a target classification network model corresponding to a current iteration period, the target article attribute is an article attribute determined from article attributes to be selected based on article attribute information to be selected and an actual breakage level corresponding to a sample article in the current iteration period;
And the information prediction module is used for carrying out classification prediction on the damage degree of the returned goods based on the target classification network model and the target goods attribute information corresponding to the returned goods so as to obtain the target damage grade corresponding to the returned goods.
In a third aspect, an embodiment of the present invention further provides an electronic device, including:
one or more processors;
A memory for storing one or more programs;
The one or more programs, when executed by the one or more processors, cause the one or more processors to implement the information prediction method as provided by any embodiment of the present invention.
In a fourth aspect, embodiments of the present invention also provide a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements an information prediction method as provided by any of the embodiments of the present invention.
One embodiment of the above invention has the following advantages or benefits:
The method comprises the steps of determining target article attributes capable of influencing article breakage degree from the article attributes to be selected based on article attribute information to be selected and actual breakage level corresponding to sample articles in a current iteration period in advance, and training based on the target article attribute information and actual breakage level corresponding to the sample articles to obtain a target classification network model capable of accurately predicting information in the current iteration period. Based on the target classification network model and the target article attribute information corresponding to the returned articles in the warehouse, the damage degree of the returned articles is classified and predicted, and the accurate target damage grade can be automatically obtained, so that the automatic prediction and the accurate prediction of the damage grade of the returned articles can be realized by using the periodically iterated target article attribute and the target classification network model, the manual participation is not needed, and the efficiency and the accuracy of the damage grade prediction are improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief description will be given below of the drawings required for the embodiments or the prior art descriptions, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for predicting information provided in one embodiment of the present invention;
FIG. 2 is a flow chart of another information prediction method provided by an embodiment of the present invention;
FIG. 3 is an example network architecture of a composite decision tree in accordance with an embodiment of the present invention;
FIG. 4 is a flow chart of yet another information prediction method provided by an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an information prediction apparatus according to an embodiment of the present invention;
Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting thereof. It should be further noted that, for convenience of description, only some, but not all of the structures related to the present invention are shown in the drawings.
Fig. 1 is a flowchart of an information prediction method according to an embodiment of the present invention, where the embodiment is applicable to predicting a breakage level of a returned article in a warehouse. The method may be performed by an information prediction device, which may be implemented in software and/or hardware, integrated in an electronic device. As shown in fig. 1, the method specifically includes the following steps:
s110, acquiring return goods in the warehouse.
The returned goods can be any goods which are successfully returned after the user purchases from the e-commerce platform. For example, a user may apply for returns in an e-commerce platform when he wants to return purchased items. The electronic commerce platform can audit the applied return description information, and if the return condition is met, the audit is confirmed to pass, and the user is allowed to return. The logistics mail the return items to the reverse warehouse of the e-commerce platform. The reverse warehouse unpacks and identifies the received returned goods, and determines whether the returned goods meet the preset returned goods conditions, such as whether the returned goods are consistent with the user description, whether the returned goods are in missing parts or other damage conditions, and the like. And if the returned goods meet the preset returned goods conditions, allowing the returned goods to be received and placed in a warehouse, and carrying out refund treatment on the user, so that the closed loop of the goods returning process is realized.
Specifically, the returned goods in the warehouse, the damage degree of which needs to be predicted, can be periodically acquired, or the returned goods stored in the warehouse in a warehouse can be acquired in real time based on the goods returning operation.
S120, acquiring a target article attribute corresponding to the current iteration period and a target classification network model, wherein the target article attribute is an article attribute determined from the article attribute to be selected based on the article attribute information to be selected corresponding to the sample article in the current iteration period and the actual breakage level, and the target classification network model is obtained by training based on the article attribute information to be selected corresponding to the sample article and the actual breakage level.
The iteration period may refer to a selection update period of the item attribute and a training update period of the network model. For example, if not currently in the item promotion phase, the iteration of the item properties and network model may be performed in one cycle, 20 to 30 days. If the current item promotion stage is in, the iteration of the item attributes and the network model can be performed with a period of 5 to 10 days. The current iteration cycle may refer to the iteration cycle in which the current time is located, that is, the most recent iteration cycle. It should be noted that, due to the continuous change of the surrounding environment, such as the change of the price of the article, the price of the consumable and the labor cost with the change of time, the influence degree of each article attribute on the damage of the article in different periods is also changed, so that the selection of the article attribute and the training of the network model are required to be periodically and iteratively updated, so as to ensure the accuracy of information prediction.
The target item attribute may refer to an item attribute that has a greater influence on item breakage in the current iteration cycle, that is, there is a correlation between the target item attribute and the item breakage degree. The candidate item attributes may refer to item attributes that may be related to the extent of breakage of the item. For example, the item attributes to be selected may include, but are not limited to, package damage level, appearance damage level, accessory integrity, warranty period remaining level, shelf life remaining level, functional integrity, security level, consumable category, storage conditions, and whether or not it is a luxury item. The item attribute information to be selected may refer to a specific attribute value of the item attribute to be selected. Since the properties of the selected items are classification parameters, they need to be digitized in order to be directly input into the model. For example, table 1 gives the information assignment results of the attributes of the selected items:
TABLE 1 information assignment results for attributes of selected items
For example, the attribute of the to-be-selected article which may affect the damage degree of the article may be first selected from all the article attributes of the sample article based on the category of the sample article, and the article attribute which is not definitely affected may be removed, for example, the sample article is a food, and the attribute of the warranty period may not exist, so that the attribute of the warranty period remaining degree may be directly removed, that is, other article attributes except the warranty period remaining degree in all the article attributes may be used as the attribute of the to-be-selected article, so that the attribute selection efficiency may be further improved.
The sample item in the current iteration period may refer to a historical return item used for training in the current iteration period, for example, an actual return item in a previous iteration period may be used as a sample item in the current iteration period to perform iterative update. The sample items in the current iteration period and the return items acquired in step S110 may belong to the same item class, so as to further ensure accuracy of information prediction. The actual breakage level may be a pre-calibrated actual breakage level of the sample article.
The target classification network model may refer to a classification network model obtained by training in the current iteration cycle. The target classification network model may be a multi-classification network model for predicting a breakage level of a return item. For example, the prediction classification may be performed using a random forest, i.e. the target classification network model is a target random forest model. The damage degree of the returned goods can be divided, and a plurality of damage grades are obtained. For example, the degree of breakage may be divided into 4 breakage levels, namely, a first breakage level, a second breakage level, a third breakage level, and a fourth breakage level. The first breakage level may be a level with a breakage degree of greater than 45%, that is, a reject level. The second breakage level may refer to a level having a breakage degree of less than or equal to 45% and greater than 25%, i.e., a regrind level. The third breakage level may refer to a level of breakage of less than or equal to 25% and greater than 10%, i.e., a level of a damaged article. The fourth breakage level may refer to a level at which the breakage degree is less than or equal to 10%, i.e., a defective grade.
Specifically, the target object attribute under each iteration period can be determined from the object attribute to be selected in advance based on the object attribute information and the actual breakage level corresponding to the sample object in each iteration period, and training is performed based on the object attribute information and the actual breakage level corresponding to the sample object, so as to obtain the target classification network model under each iteration period. In practical application, the target object attribute and the target classification network model under the predetermined current iteration period can be directly obtained, so that the information prediction can be carried out by utilizing the latest target object attribute and the target classification network model, and the accuracy of the information prediction can be effectively ensured.
S130, based on the target classification network model and the target item attribute information corresponding to the returned items, classifying and predicting the damage degree of the returned items, and obtaining the target damage grade corresponding to the returned items.
The target item attribute information corresponding to the returned item may refer to a specific attribute value of the target item attribute of the returned item. The breakage level may refer to a ratio between a maintenance cost of the returned item and a value attribute value of the returned item. The maintenance cost may refer to the cost of refurbishing the returned items. The return item value attribute value may refer to an original price of the return item. The greater the degree of breakage, the higher the breakage level.
Specifically, the attribute information of the target object corresponding to the returned object can be obtained, and the attribute information of the target object obtained at this time is the assigned attribute information, so that the attribute information of the target object corresponding to the returned object can be directly input into the target classification network model to perform classification prediction of the damage degree. The target classification network model can output the predicted target damage level, so that the target classification network model iterated periodically can be utilized to automatically and accurately predict the damage level of the returned goods, and the accuracy and the efficiency of information prediction are improved.
According to the technical scheme, the target article attribute which can influence the article damage degree is determined from the article attribute to be selected in advance based on the article attribute information to be selected and the actual damage level which correspond to the sample article in the current iteration period, training is carried out based on the article attribute information to be selected and the actual damage level which correspond to the sample article, and a target classification network model which can accurately predict information in the current iteration period is obtained. Based on the target classification network model and the target article attribute information corresponding to the returned articles in the warehouse, the damage degree of the returned articles is classified and predicted, and the accurate target damage grade can be automatically obtained, so that the automatic prediction and the accurate prediction of the damage grade of the returned articles can be realized by using the periodically iterated target article attribute and the target classification network model, the manual participation is not needed, and the efficiency and the accuracy of the damage grade prediction are improved.
Fig. 2 is a flowchart of another information prediction method according to an embodiment of the present invention, where, based on the foregoing embodiments, a detailed description is given to a process of determining a target object attribute in a current iteration period, and a detailed description is given to a training process when the target classification network model is a target random forest model. Wherein the explanation of the same or corresponding terms as those of the above embodiments is not repeated herein.
Referring to fig. 2, another information prediction method provided in this embodiment specifically includes the following steps:
s210, determining a target connection function corresponding to the current iteration period from a plurality of connection functions to be selected.
The connection function to be selected may be a selectable connection function. The target connection function may refer to the most suitable connection function at the current iteration cycle. The join function may refer to a function in logistic regression that is used to convert continuous regression values to discrete class values so that a logistic regression fit to the class data may be performed. For example, the join function may include, but is not limited to, a logic function, a Probit function, a supplemental log-log function, a negative log-log function, and Cauchit functions. Different connection functions may be used to describe different dependent variable distribution situations. For example, the log function is used when the dependent variable distribution is uniform or the number of categories is small, the Probit function is used when the dependent variable is close to normal distribution, the supplementary log-log function is used when the probability of occurrence of categories with high dependent variable level is high and the number of categories is large, the negative log-log function is used when the probability of occurrence of categories with low dependent variable level is high and the number of categories is large, and the Cauchit function is used when the dependent variable has an extreme value.
Specifically, the most suitable target connection function in the current iteration period can be determined from a plurality of connection functions to be selected based on the use range of each connection function to be selected and the distribution condition of the actual damage level corresponding to the sample object in the current iteration period.
Illustratively, the step S210 may include checking a logistic regression model corresponding to each candidate connection function based on a preset hypothesis test mode, determining a hypothesis test parameter value corresponding to each candidate connection function, determining candidate connection functions from a plurality of candidate connection functions based on the hypothesis test parameter value, and determining an objective connection function corresponding to a current iteration period from the candidate connection functions based on distribution information of actual breakage levels corresponding to sample articles.
Specifically, a corresponding logistic regression model may be constructed based on each connection function to be selected, and based on a preset hypothesis test mode, such as pearson test or De-viance test mode, a hypothesis test may be performed on each constructed logistic regression model to obtain a hypothesis test parameter value corresponding to each logistic regression model, i.e., a P value, so as to obtain a hypothesis test parameter value corresponding to each connection function to be selected. The candidate connection function whose hypothesis test parameter value is less than or equal to 0.05 may be determined as a candidate connection function, that is, the candidate connection function having a significant relationship for which the hypothesis holds is determined as a candidate connection function. After determining the candidate connection functions, a best-matching target connection function may be determined from the candidate connection functions based on the usage range of each candidate connection function and the distribution information of the actual breakage level corresponding to the sample article in the current iteration cycle. By checking the logistic regression model corresponding to the connection function to be selected, a more matched target connection function can be determined, and the accuracy of attribute selection is further improved.
S220, performing multiple ordered logistic regression based on the target connection function, the attribute information of the selected object corresponding to the sample object in the current iteration period and the actual breakage level, and determining regression index information corresponding to each object attribute.
The multiple ordered logistic regression is a logistic regression mode adopted when the dependent variable (namely the breakage level) is multi-classification data and the numerical values have a sequential relationship. The regression index information may be index information for measuring the degree of influence of the attribute of the selected item on the breakage of the item. Regression index information may also be used to characterize the importance of the attributes of the selected item. For example, the regression index information may include at least one of a regression coefficient, a hypothesis test parameter value (i.e., P value), and an attribute variation amplitude index value (i.e., OR value). The regression coefficient can be used for representing the influence degree of the attribute of the selected article on the damage of the article, and the larger the regression coefficient is, the larger the influence degree is. The assumption that the value of the inspection parameter can be used to characterize how significantly the attribute of the selected item affects breakage of the item, the smaller the value of the inspection parameter, the more significant the effect. The attribute change amplitude index value refers to the change amplitude of the dependent variable after the independent variable is increased by one unit. An attribute change magnitude index value greater than 1 indicates that the dependent variable increases with increasing independent variable, whereas it decreases with increasing independent variable.
Specifically, the multiple ordered logistic regression model may be expressed as follows:
Where δ 1 is a constant term, k is the number of breakage levels, such as k equals 4, pi 1～πk-1 is the probability of 1 to k-1, p k-1 is the ratio of the occurrence to non-occurrence of the k-1 class of results, pi 1+…+πk-1 is the cumulative probability of the occurrence of the k=1 to k=k-1 class of results, X 1 to X m are candidate item attribute information, β 1 to β m are regression coefficients, OR is used to measure the magnitude of the dependent variable lifting when the independent variable changes by one unit.
Through the multiple ordered logistic regression model, multiple ordered logistic regression is performed based on the target connection function, the attribute information of the selected object corresponding to the sample object in the current iteration period and the actual damage level, so that the attribute information of the selected object and the actual damage level can be fitted, and regression index information corresponding to each attribute of the selected object in the multiple ordered logistic regression model after final fitting is obtained.
S230, determining the object attribute from the object attributes to be selected based on the regression index information.
Specifically, regression index information corresponding to each item attribute to be selected may be compared with an index information threshold, and the item attribute to be selected having a significant relationship with the degree of loss may be determined as the target item attribute. For example, the candidate item attributes may be determined from the candidate item attributes based on the hypothesis test parameter values corresponding to each candidate item attribute, and the target item attributes may be determined from the candidate item attributes based on the regression coefficients or the attribute variation amplitude index values corresponding to each candidate item attribute. For example, an item attribute to be selected whose hypothesis test parameter value is smaller than a first threshold value (such as 0.05) may be determined as a candidate item attribute, or an item attribute to be selected whose hypothesis test parameter value is smaller than the first threshold value (such as 0.05) and larger than a second threshold value (such as 0.01) may be determined as a candidate item attribute, and an item attribute to be selected whose hypothesis test parameter value is smaller than the second threshold value (such as 0.01) may be directly determined as a target item attribute, so that an item attribute to be selected having a significant relationship may be directly determined as a target item attribute based on the hypothesis test parameter value. And determining the candidate object attribute with the regression coefficient or the attribute variation amplitude index value larger than the preset threshold value as the object attribute, so that the object attribute which is obviously related to the object damage can be accurately determined by using the regression index information.
S240, sampling object attributes corresponding to the sample objects, and determining sample data corresponding to each composite decision tree, wherein the sample data comprise part of object attribute information and actual breakage levels corresponding to the sample objects, and each composite decision tree comprises at least two decision trees with hierarchical structures.
The random forest model can comprise a plurality of composite decision trees, the network structure of each composite decision tree is the same, but the sample data used in the training of each composite decision tree is different, so that a plurality of different composite decision trees can be trained. A composite decision tree is a network model for multi-categorizing the degree of damage to an item. Typically a decision tree is a two-class network model, so that at least two decision trees need to be connected to achieve multi-class prediction. For example, FIG. 3 shows an example of a network architecture for a composite decision tree. The degree of breakage is classified into 4 breakage levels, namely, a first breakage level at which the degree of breakage F is greater than 45%, a second breakage level at which the degree of breakage F is less than or equal to 45% and greater than 25%, a third breakage level at which the degree of breakage F is less than or equal to 25% and greater than 10%, and a fourth breakage level at which the degree of breakage F is less than or equal to 10%. The composite decision tree may include 3 decision trees having a hierarchical structure, tree 1, tree 2, and tree 3, respectively. The tree 1 is located at the top layer and used for performing two classifications with the breakage degree F being greater than 25% and F being less than or equal to 25%, the tree 2 and the tree 3 are located at the bottom layer, and the tree 2 is used for performing two classifications with the breakage degree F being greater than 45% and F being less than or equal to 45%, so that the classification of the first breakage level and the second breakage level is achieved. The tree 3 is used for classifying the degree of breakage F greater than 10% and F less than or equal to 10%, thereby achieving classification of the third breakage level and the fourth breakage level. By taking the leaf node of the tree 1 as the root node of the tree 2 and the tree 3, the combination of a plurality of decision trees is realized, and a composite decision tree for realizing multi-classification is constructed.
Specifically, a composite decision tree having a hierarchical structure is constructed based on the number of damage levels and the damage level demarcation value corresponding to each damage level. The number of the composite decision trees is a plurality, and the specific number of the composite decision trees can be determined based on service requirements. Based on the number of the composite decision trees, the object attributes corresponding to the sample objects can be sampled randomly, part of object attributes sampled for each composite decision tree are determined, and part of object attribute information and actual breakage level are used as sample data of the corresponding composite decision tree.
S250, training each composite decision tree based on sample data, and constructing a target random forest model based on the trained composite decision tree.
In particular, the training of each composite decision tree may be independent of each other. Training each decision tree in the composite decision tree based on the part of article attribute information and the actual damage level corresponding to each composite decision tree, so as to obtain a composite decision tree after training is finished, combining a plurality of composite decision trees after training is finished, and obtaining a target random forest model after training in the current iteration period.
The training of each composite decision tree based on the sample data in S250 may include, for each composite decision tree, processing sample data corresponding to each composite decision tree based on a classification result of each decision tree in the composite decision tree to obtain sub-sample data corresponding to each decision tree in the composite decision tree, where the sub-sample data includes part of item attribute information and damage level labels corresponding to sample items, and training each decision tree based on the sub-sample data to obtain a trained composite decision tree.
Specifically, for each composite decision tree, based on a classification result (i.e., a damage level that each decision tree can classify) corresponding to each decision tree in the composite decision tree, sample data corresponding to the composite decision tree is selected and an actual damage level is adjusted, so as to obtain a damage level label and corresponding partial article attribute information corresponding to a sample article that each decision tree can predict. For example, referring to fig. 2, for tree 1, all sample articles corresponding to the composite decision tree may be taken as sample articles of tree 1, and the actual breakage level corresponding to each sample article may be adjusted to be a breakage level label matched with the sample article, for example, when the actual breakage level is a second breakage level with a breakage level F greater than 45%, the adjusted breakage level label is a breakage level with a breakage level F greater than 25%. A sample article corresponding to a breakage level with a breakage degree F of greater than 25% may be used as a sample article of the tree 2, and a sample article corresponding to a breakage level with a breakage degree F of less than or equal to 25% may be used as a sample article of the tree 3. Since trees 2 and 3 are the lowest level decision trees, the actual breakage level corresponding to the sample item can be directly determined as the breakage level label. Similarly, sub-sample data corresponding to each decision tree in the composite decision tree may be obtained. The training process for each decision tree in the composite decision tree is also the same. For example, the training process of each decision tree may be that a node is generated based on sub-sample data, and whether the node is a leaf node is detected, if not, an attribute is divided for the current node, a branch is generated by dividing the current node, whether sample data exists in each branch is judged, if the sample data exists, the node is continuously generated based on the sample data, if the sample data does not exist, the branch is converted into the leaf node, and the decision tree is constructed until the generated node is the leaf node.
The leaf node is the node for finally judging the class of the damaged grade, if the data of the current node all belong to one damaged grade, the current node is the leaf node, and the label of the current node is the damaged grade label of the data of the current node. To prevent infinite iterations, the maximum depth of the decision tree may be set, if the constructed decision tree is high to the maximum depth, the current node is a leaf node, and the majority of labels in the current node are leaf node labels.
With the continuous division of data, the purity of the data in the nodes is higher and higher, and the data can be divided by selecting the optimal division attribute from the attribute set. The embodiment can judge the optimal attribute by taking the information entropy gain rate as a label. The information entropy is an index for measuring the purity of data. The entropy of information is inversely related to the purity of the data set. Assuming that the current dataset is D, where the i-th class of data is p i (i=1, 2,.|γ|), the entropy of information of the current dataset D is:
The information gain is a measure of the amount of purity improvement that the current data set is partitioned by attribute a. Assuming that attribute a has m values, partitioning the data set with this attribute results in m subsets (D 1,D2,…Dm). The information entropy of each subset can be obtained based on the above formula. The information entropy of each subset is weighted according to the data quantity of the subset, so that the information gain of the current data set D is obtained by dividing the attribute a, namely:
wherein, the larger the information gain, the greater the improvement in data purity by attribute a partitioning. The information entropy gain rate can be expressed as:
the denominator of the equation increases with the number of subsets m, so that the information gain can be balanced with the information entropy gain rate.
After determining the optimal partitioning attribute of the current node, assuming that the current attribute has m values, partitioning the dataset D with the attribute may obtain m branches (D 1,D2,…Dm). For each branch, if the current branch has no sample data, the branch is a leaf node, the label of the branch is the label with the largest number in D, otherwise, the branch is used as a new node to continue to generate a tree. It should be noted that, because the classification prediction is performed by adopting a random forest mode, the front pruning or the rear pruning of the decision tree is not required, and the generalization of the classification prediction can be enhanced.
S260, acquiring returned goods in the warehouse.
S270, based on the target random forest model and the target item attribute information corresponding to the returned items, classifying and predicting the damage degree of the returned items, and obtaining the target damage grade corresponding to the returned items.
Specifically, the attribute information of the target object corresponding to the returned object can be input into the target random forest model, each composite decision tree in the target random forest model carries out classification prediction of damage degree based on the input attribute information of the target object, the predicted damage level output by each composite decision tree is obtained, and the predicted damage level with the highest occurrence number in all the predicted damage levels is determined to be the final target damage level, so that automatic prediction and accurate prediction of the damage level of the returned object can be realized by utilizing the target random forest model, and the efficiency and accuracy of damage level prediction are improved.
According to the technical scheme, multiple ordered logistic regression is performed based on the target connection function, the attribute information of the selected object corresponding to the sample object in the current iteration period and the actual damage level, and the target object attribute can be accurately determined from the object attributes based on the regression index information corresponding to each object attribute to be selected, so that the accuracy of object attribute selection is improved. By constructing a composite decision tree comprising at least two decision trees with a hierarchical structure and constructing a target random forest model based on a plurality of composite decision trees, multi-classification prediction of the target random forest model can be realized, and the accuracy of classification prediction is ensured.
Fig. 4 is a flowchart of another information prediction method according to an embodiment of the present invention, where processing operations for performing a home right transfer on return items are added on the basis of the above embodiments. Wherein the explanation of the same or corresponding terms as those of the above embodiments is not repeated herein.
Referring to fig. 4, another information prediction method provided in this embodiment specifically includes the following steps:
s410, acquiring returned goods in the warehouse.
S420, acquiring a target article attribute corresponding to the current iteration period and a target classification network model, wherein the target article attribute is an article attribute determined from the article attribute to be selected based on the article attribute information to be selected corresponding to the sample article in the current iteration period and the actual breakage level, and the target classification network model is obtained by training based on the article attribute information to be selected corresponding to the sample article and the actual breakage level.
S430, based on the target classification network model and the target item attribute information corresponding to the returned items, classifying and predicting the damage degree of the returned items, and obtaining the target damage grade corresponding to the returned items.
S440, performing attribution right transfer processing on the returned goods based on the target breakage level, and generating attribution right transfer tasks corresponding to the returned goods.
The attribute right transferring process may refer to that the e-commerce platform transfers the attribute right of the returned goods to other objects except the goods supplier. For example, the ownership transfer process may refer to a resale process of return items. The ownership transfer task may be a task of transferring ownership of the return item to an object other than the item provider. For example, the ownership transfer task may refer to an order after a resale, i.e., a purchase order for a return item.
Specifically, the target damage level corresponding to the returned goods can be displayed on the e-commerce platform, so that the user interested in the returned goods can purchase the returned goods, the attribution right transfer task corresponding to the returned goods can be generated, and attribution right transfer of the returned goods is realized. The electronic commerce platform can realize secondary selling of the returned goods without returning the returned goods to the goods supplier, so that the aim of reducing damage and creating income can be fulfilled, the risk of secondary damage generated in the process of returning the returned goods to the supplier is avoided, and the viscosity of the electronic commerce platform to the supplier is further increased.
Illustratively, S440 may include determining a target value attribute value corresponding to the returned item based on the target breakage level, and displaying target item information corresponding to the returned item, the target item information including the target value attribute value and the target breakage level, and generating an attribution transfer task corresponding to the returned item based on the object information of the target object in response to an item acquisition operation triggered by the target object for the displayed target item information.
The target value attribute value may be a price at the time of resale of the return item. The smaller the target breakage level, the greater the degree of breakage and the lower the corresponding target value attribute value. The target object may refer to an object that has the right to return goods after resale. The target object may be an individual user or an enterprise, etc. The object information of the target object may include contact information and a receiving address of the target object, and the like.
Specifically, the original value attribute value corresponding to the returned item may be processed based on the target breakage level to obtain a target value attribute value that is less than the original value attribute value. Target item information including a target value attribute value and a target breakage level may be presented in a second hand resale platform in the e-commerce platform. If the target object wants to purchase the returned goods based on the displayed target goods information, the target object can trigger the goods acquisition operation, such as clicking a purchase button, so that a attribution right transfer task corresponding to the returned goods can be generated based on the object information of the target object, secondary selling of the returned goods is realized, the returned goods is not required to be returned to the goods supplier, damage prevention and cost reduction of the supplier can be realized, and the revenue generation of the electronic commerce platform is increased.
For example, the e-commerce platform may send the target breakage level, the target value attribute value, and the grading video of the returned item corresponding to the returned item to an item provider of the returned item to support the item provider to audit to determine whether to allow the e-commerce platform to transfer the ownership. If the verification is passed, the return goods are subjected to attribution right transfer processing, otherwise, the return goods are returned to the goods supplier, so that the personalized requirements of the goods supplier are met.
According to the technical scheme, the return goods are subjected to the attribution right transfer processing based on the target damage level, and the attribution right transfer task corresponding to the return goods is generated, so that the secondary selling of the return goods can be realized, the return goods are not required to be returned to the goods supplier, the risk of secondary damage in the return process of the return goods to the supplier can be avoided, and the viscosity of the electronic commerce platform to the supplier is further increased.
Based on the above technical solution, step S440 may include detecting whether the returned goods meet the preset attribution right transfer condition based on attribution right transfer configuration information of the goods supplier corresponding to the returned goods, and if the returned goods meet the preset attribution right transfer condition, performing attribution right transfer processing on the returned goods based on the target breakage level, to generate an attribution right transfer task corresponding to the returned goods.
Wherein the ownership transfer configuration information may be used to characterize the item information that the provider allows for transfer of ownership. For example, the assignment transfer configuration may be performed in the category dimension or the vendor dimension. The ownership transfer configuration information may include at least one category in which the provider allows transfer of ownership.
Specifically, the home right transfer configuration information of each item provider may be preconfigured on the e-commerce platform. In practical application, the attribution right transfer configuration information of the article supplier corresponding to the returned article can be obtained, whether the returned article meets the preset attribution right transfer condition or not is detected based on the attribution right transfer configuration information, for example, whether the article class of the returned article belongs to the configured article class is detected, if yes, the attribution right transfer of the returned article is allowed, at the moment, attribution right transfer processing is carried out on the returned article based on the target damage grade, otherwise, the returned article is returned to the article supplier, thereby realizing the configurability of attribution right transfer and meeting the resale demands of different articles of the supplier.
The following is an embodiment of an information prediction apparatus provided in the present embodiment, which belongs to the same inventive concept as the information prediction method of the above embodiments, and reference may be made to the embodiment of the information prediction method for details that are not described in detail in the embodiment of the information prediction apparatus.
Fig. 5 is a schematic structural diagram of an information prediction apparatus according to an embodiment of the present invention, which is applicable to predicting damage of returned goods in a warehouse. As shown in FIG. 5, the apparatus specifically includes a return item acquisition module 510, a model acquisition module 520, and an information prediction module 530.
The system comprises a return goods obtaining module 510, a model obtaining module 520 and an information predicting module 530, wherein the return goods obtaining module is used for obtaining a return goods in a warehouse, the model obtaining module 520 is used for obtaining a target goods attribute corresponding to a current iteration period and a target classification network model, the target goods attribute is an article attribute determined from the selected goods attribute based on the to-be-selected article attribute information and the actual damage level corresponding to a sample article in the current iteration period, the target classification network model is obtained by training based on the target goods attribute information and the actual damage level corresponding to the sample article, and the information predicting module 530 is used for carrying out classification prediction on the damage degree of the return goods based on the target classification network model and the target goods attribute information corresponding to the return goods to obtain the target damage level corresponding to the return goods.
According to the technical scheme, the target article attribute which can influence the article damage degree is determined from the article attribute to be selected in advance based on the article attribute information to be selected and the actual damage level which correspond to the sample article in the current iteration period, training is carried out based on the article attribute information to be selected and the actual damage level which correspond to the sample article, and a target classification network model which can accurately predict information in the current iteration period is obtained. Based on the target classification network model and the target article attribute information corresponding to the returned articles in the warehouse, the damage degree of the returned articles is classified and predicted, and the accurate target damage grade can be automatically obtained, so that the automatic prediction and the accurate prediction of the damage grade of the returned articles can be realized by using the periodically iterated target article attribute and the target classification network model, the manual participation is not needed, and the efficiency and the accuracy of the damage grade prediction are improved.
Optionally, the apparatus further comprises:
The target connection function determining module is used for determining a target connection function corresponding to the current iteration period from a plurality of connection functions to be selected;
The regression index information determining module is used for performing multiple ordered logistic regression based on the target connection function, the attribute information of the selected object corresponding to the sample object in the current iteration period and the actual damage level, and determining regression index information corresponding to each object attribute;
and the target object attribute determining module is used for determining the target object attribute from the object attributes to be selected based on the regression index information.
Optionally, the objective connection function determining module is specifically configured to:
The method comprises the steps of detecting a logistic regression model corresponding to each connection function to be selected based on a preset hypothesis detection mode, determining a hypothesis detection parameter value corresponding to each connection function to be selected, determining candidate connection functions from a plurality of connection functions to be selected based on the hypothesis detection parameter value, and determining a target connection function corresponding to a current iteration period from the candidate connection functions based on distribution information of actual breakage levels corresponding to sample articles.
Optionally, the regression indicator information includes at least one of a regression coefficient, a hypothesis test parameter value, and an attribute variation amplitude indicator value.
Optionally, the target classification network model is a target random forest model, and the device further comprises:
The sample data determining module is used for sampling the object attribute corresponding to the sample object and determining sample data corresponding to each composite decision tree, wherein the sample data comprises partial object attribute information corresponding to the sample object and actual breakage level, and each composite decision tree comprises at least two decision trees with hierarchical structures;
and the composite decision tree training module is used for training each composite decision tree based on the sample data and constructing a target random forest model based on the trained composite decision tree.
Optionally, the composite decision tree training module is specifically configured to:
Aiming at each composite decision tree, based on the classification result of each decision tree in the composite decision tree, sample data corresponding to the composite decision tree is processed to obtain sub-sample data corresponding to each decision tree in the composite decision tree, wherein the sub-sample data comprises part of article attribute information and damage grade labels corresponding to sample articles, and training is carried out on each decision tree based on the sub-sample data to obtain the trained composite decision tree.
Optionally, the apparatus further comprises:
And the attribution right transfer module is used for carrying out attribution right transfer processing on the returned goods based on the target damage grade after obtaining the target damage grade corresponding to the returned goods, and generating attribution right transfer tasks corresponding to the returned goods.
Optionally, the attribution right transferring module is specifically configured to:
Determining a target value attribute value corresponding to the returned goods based on the target damage level, and displaying target goods information corresponding to the returned goods, wherein the target goods information comprises the target value attribute value and the target damage level;
and responding to an article acquisition operation triggered by a target object aiming at the displayed target article information, and generating a attribution right transfer task corresponding to the returned article based on the object information of the target object.
Optionally, the attribution right transferring module is specifically configured to:
Detecting whether the returned goods meet preset attribution right transfer conditions or not based on attribution right transfer configuration information of an goods supplier corresponding to the returned goods, and if the returned goods meet the preset attribution right transfer conditions, carrying out attribution right transfer processing on the returned goods based on the target breakage level to generate attribution right transfer tasks corresponding to the returned goods.
The information prediction device provided by the embodiment of the invention can execute the information prediction method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of executing the information prediction method.
It should be noted that, in the embodiment of the information prediction apparatus, each unit and module included are only divided according to the functional logic, but not limited to the above division, so long as the corresponding functions can be implemented, and the specific names of the functional units are only for convenience of distinguishing each other, and are not used to limit the protection scope of the present invention.
Fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention. Fig. 6 illustrates a block diagram of an exemplary electronic device 12 suitable for use in implementing embodiments of the present invention. The electronic device 12 shown in fig. 6 is merely an example and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.
As shown in fig. 6, the electronic device 12 is in the form of a general purpose computing device. The components of the electronic device 12 may include, but are not limited to, one or more processors or processing units 16, a system memory 28, and a bus 18 that connects the various system components, including the system memory 28 and the processing units 16.
Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, and a local bus using any of a variety of bus architectures. By way of example, and not limitation, such architectures include Industry Standard Architecture (ISA) bus, micro channel architecture (MAC) bus, enhanced ISA bus, video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.
Electronic device 12 typically includes a variety of computer system readable media. Such media can be any available media that is accessible by electronic device 12 and includes both volatile and nonvolatile media, removable and non-removable media.
The system memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM) 30 and/or cache memory 32. The electronic device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read from or write to non-removable, nonvolatile magnetic media (not shown in FIG. 6, commonly referred to as a "hard disk drive"). Although not shown in fig. 6, a magnetic disk drive for reading from and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), and an optical disk drive for reading from or writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In such cases, each drive may be coupled to bus 18 through one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to carry out the functions of the embodiments of the invention.
A program/utility 40 having a set (at least one) of program modules 42 may be stored in, for example, system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment. Program modules 42 generally perform the functions and/or methods of the embodiments described herein.
The electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), one or more devices that enable a user to interact with the electronic device 12, and/or any devices (e.g., network card, modem, etc.) that enable the electronic device 12 to communicate with one or more other computing devices. Such communication may occur through an input/output (I/O) interface 22. Also, the electronic device 12 may communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN) and/or a public network, such as the Internet, through a network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 over the bus 18. It should be appreciated that although not shown, other hardware and/or software modules may be used in connection with electronic device 12, including, but not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data backup storage systems, and the like.
The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, for example, implementing an information prediction method step provided by any embodiment of the present invention, the method comprising:
The method comprises the steps of obtaining a returned article in a warehouse, obtaining a target article attribute corresponding to a current iteration period and a target classification network model, wherein the target article attribute is an article attribute determined from the article attribute to be selected based on article attribute information to be selected corresponding to a sample article in the current iteration period and an actual breakage level, the target classification network model is obtained by training based on the target article attribute information corresponding to the sample article and the actual breakage level, and the returned article is subjected to classification prediction of breakage degree based on the target classification network model and the target article attribute information corresponding to the returned article to obtain the target breakage level corresponding to the returned article.
Of course, those skilled in the art will understand that the processor may also implement the technical solution of the information prediction method provided in any embodiment of the present invention.
The present embodiment provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the information prediction method provided by any of the embodiments of the present invention.
The computer storage media of embodiments of the invention may take the form of any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium can be, for example but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, either in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations of the present invention may be written in one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (for example, through the Internet using an Internet service provider).
It will be appreciated by those of ordinary skill in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be centralized on a single computing device, or distributed over a network of computing devices, or they may alternatively be implemented in program code executable by a computer device, such that they are stored in a memory device and executed by the computing device, or they may be separately fabricated as individual integrated circuit modules, or multiple modules or steps within them may be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any specific combination of hardware and software.
Note that the above is only a preferred embodiment of the present invention and the technical principle applied. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, while the invention has been described in connection with the above embodiments, the invention is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the invention, which is set forth in the following claims.
Claims (12)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311405589.9A CN119904166A (en) | 2023-10-26 | 2023-10-26 | Information prediction method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311405589.9A CN119904166A (en) | 2023-10-26 | 2023-10-26 | Information prediction method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119904166A true CN119904166A (en) | 2025-04-29 |
Family
ID=95466949
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311405589.9A Pending CN119904166A (en) | 2023-10-26 | 2023-10-26 | Information prediction method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119904166A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019242286A1 (en) * | 2018-06-19 | 2019-12-26 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for allocating service requests |
| US20200349503A1 (en) * | 2019-04-30 | 2020-11-05 | Truenorth Systems Ltd | Method and system for predicting extent of loss |
| CN111967632A (en) * | 2020-04-07 | 2020-11-20 | 上海寻梦信息技术有限公司 | Method and device for determining packaging grade of logistics object, electronic equipment and storage medium |
| CN113642800A (en) * | 2021-08-20 | 2021-11-12 | 林周县众陶联供应链服务有限公司 | Data analysis method and data analysis system for firing system of architectural ceramic kiln |
| CN114331284A (en) * | 2021-12-31 | 2022-04-12 | 深圳市凯东源现代物流股份有限公司 | Intelligent warehousing service management system based on cloud computing |
| US20220156682A1 (en) * | 2020-11-18 | 2022-05-19 | Cisco Technology, Inc. | Determining the shared fate of damaged items within a supply chain |
| CN114912798A (en) * | 2022-05-16 | 2022-08-16 | 中震科建(广东)防灾减灾研究院有限公司 | Earthquake loss evaluation system based on random forest and earthquake damage big data |
-
2023
- 2023-10-26 CN CN202311405589.9A patent/CN119904166A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019242286A1 (en) * | 2018-06-19 | 2019-12-26 | Beijing Didi Infinity Technology And Development Co., Ltd. | Systems and methods for allocating service requests |
| US20200349503A1 (en) * | 2019-04-30 | 2020-11-05 | Truenorth Systems Ltd | Method and system for predicting extent of loss |
| CN111967632A (en) * | 2020-04-07 | 2020-11-20 | 上海寻梦信息技术有限公司 | Method and device for determining packaging grade of logistics object, electronic equipment and storage medium |
| US20220156682A1 (en) * | 2020-11-18 | 2022-05-19 | Cisco Technology, Inc. | Determining the shared fate of damaged items within a supply chain |
| CN113642800A (en) * | 2021-08-20 | 2021-11-12 | 林周县众陶联供应链服务有限公司 | Data analysis method and data analysis system for firing system of architectural ceramic kiln |
| CN114331284A (en) * | 2021-12-31 | 2022-04-12 | 深圳市凯东源现代物流股份有限公司 | Intelligent warehousing service management system based on cloud computing |
| CN114912798A (en) * | 2022-05-16 | 2022-08-16 | 中震科建(广东)防灾减灾研究院有限公司 | Earthquake loss evaluation system based on random forest and earthquake damage big data |
Non-Patent Citations (3)
| Title |
|---|
| SATVIK GARG ET AL.: "Towards a Multimodal System for Precision Agriculture using IoT and Machine Learning", 《12TH ICCCNT 2021》, 3 November 2021 (2021-11-03) * |
| 袁炜: "DS公司进口食品添加剂库存管理研究", 《中国优秀硕士学位论文全文数据库 (经济与管理科学辑)》, no. 11, 15 November 2018 (2018-11-15), pages 152 - 322 * |
| 郭海峰;林群;: "不确定条件下多产品第四方物流闭环网络鲁棒优化研究", 井冈山大学学报(自然科学版), vol. 28, no. 03, 15 May 2018 (2018-05-15), pages 38 - 44 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9064212B2 (en) | Automatic event categorization for event ticket network systems | |
| CN113778979B (en) | A method and device for determining click rate of live broadcast | |
| CN111090822A (en) | Method and device for pushing business objects | |
| CN113393306A (en) | Product recommendation method and device, electronic equipment and computer readable medium | |
| US10678821B2 (en) | Evaluating theses using tree structures | |
| CN113159355A (en) | Data prediction method, data prediction device, logistics cargo quantity prediction method, medium and equipment | |
| CN113743971A (en) | Data processing method and device | |
| CN110688536A (en) | Label prediction method, device, equipment and storage medium | |
| CN115423575B (en) | Internet-based digital analysis management system and method | |
| CN119090542B (en) | An e-commerce platform order analysis method and system based on data mining | |
| US20130325863A1 (en) | Data Clustering for Multi-Layer Social Link Analysis | |
| Zhou et al. | An advanced inventory data mining system for business intelligence | |
| CN113869596B (en) | Task prediction processing method, device, product and medium | |
| CN113159877B (en) | Data processing method, device, system, and computer-readable storage medium | |
| US20210142348A1 (en) | Multi-layered system for heterogeneous pricing decisions by continuously learning market and hotel dynamics | |
| CN116362790A (en) | Client type prediction method, client type prediction device, electronic device, medium and program product | |
| CN110348922B (en) | Method and apparatus for generating information | |
| CN119760202B (en) | Task execution method and device, electronic device, storage medium and program product | |
| CN119904166A (en) | Information prediction method, device, equipment and storage medium | |
| CN110503482B (en) | Article processing method, device, terminal and storage medium | |
| CN112785234A (en) | Goods recommendation method, device, equipment and storage medium | |
| CN110490682B (en) | Method and device for analyzing commodity attributes | |
| CN110334177B (en) | Semantic similarity model training and semantic similarity recognition methods and devices and electronic equipment | |
| CN114548538B (en) | Task volume prediction method, device, electronic device and storage medium | |
| CN113554460B (en) | A method and device for identifying potential users |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |