CN116579803A - Multi-class joint demand prediction method and device based on substitution and association effects - Google Patents
Multi-class joint demand prediction method and device based on substitution and association effects Download PDFInfo
- Publication number
- CN116579803A CN116579803A CN202310380958.7A CN202310380958A CN116579803A CN 116579803 A CN116579803 A CN 116579803A CN 202310380958 A CN202310380958 A CN 202310380958A CN 116579803 A CN116579803 A CN 116579803A
- Authority
- CN
- China
- Prior art keywords
- commodity
- association
- attribute
- substitution
- commodities
- 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
- 238000006467 substitution reaction Methods 0.000 title claims abstract description 99
- 238000000034 method Methods 0.000 title claims abstract description 68
- 230000000694 effects Effects 0.000 title claims abstract description 23
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 32
- 238000007476 Maximum Likelihood Methods 0.000 claims abstract description 17
- 238000005065 mining Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 6
- 238000007637 random forest analysis Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000000737 periodic effect Effects 0.000 claims description 3
- 230000006399 behavior Effects 0.000 abstract description 16
- 238000010801 machine learning Methods 0.000 abstract description 6
- 238000007405 data analysis Methods 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 5
- 230000035772 mutation Effects 0.000 description 4
- 238000003066 decision tree Methods 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001737 promoting effect Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- MXBCYQUALCBQIJ-RYVPXURESA-N (8s,9s,10r,13s,14s,17r)-13-ethyl-17-ethynyl-11-methylidene-1,2,3,6,7,8,9,10,12,14,15,16-dodecahydrocyclopenta[a]phenanthren-17-ol;(8r,9s,13s,14s,17r)-17-ethynyl-13-methyl-7,8,9,11,12,14,15,16-octahydro-6h-cyclopenta[a]phenanthrene-3,17-diol Chemical compound OC1=CC=C2[C@H]3CC[C@](C)([C@](CC4)(O)C#C)[C@@H]4[C@@H]3CCC2=C1.C1CC[C@@H]2[C@H]3C(=C)C[C@](CC)([C@](CC4)(O)C#C)[C@@H]4[C@@H]3CCC2=C1 MXBCYQUALCBQIJ-RYVPXURESA-N 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000015895 biscuits Nutrition 0.000 description 1
- 239000008370 chocolate flavor Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000009396 hybridization Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 235000015067 sauces Nutrition 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 238000013526 transfer learning Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Accounting & Taxation (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Entrepreneurship & Innovation (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Genetics & Genomics (AREA)
- Physiology (AREA)
- General Business, Economics & Management (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- Probability & Statistics with Applications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a multi-class combined demand prediction method and a device based on substitution and association, wherein the prediction method comprises the following steps: determining core attributes of the commodities based on sales information of the commodities currently being sold; defining substitution and association actions among all attribute combination commodities in the purchasing behavior of consumers; and incorporating parameters for measuring the replacement and the association into an operation planning model established based on the maximum likelihood idea, and solving the actual demand and the demand proportion of all attribute combined commodities comprising the existing commodity and the new commodity by using a meta heuristic algorithm. The prediction method of the invention can be applied to the actual demand prediction of the multi-class goods of the retail store, the demand prediction is jointly enabled based on the data analysis capability of machine learning and the planning capability of operation planning technology, the actual demand of the goods is determined on the basis of considering the substitution and association effects in the purchasing behavior of consumers, and the demand prediction model which is usually used for single class is expanded to be applicable to the situation of two/more associated classes.
Description
Technical Field
The invention relates to the technical field of data, in particular to a multi-class joint demand prediction method and device based on substitution and association.
Background
The commodity demand prediction includes prediction of existing commodity and prediction of new commodity. The sales performance prediction of the new commodity has key influence on the successful release of the commodity, can help companies develop markets, and improves the accuracy and instantaneity of the new demand prediction; secondly, the method is also helpful to solve challenges brought by rapid iteration commodity combination in a new retail background for class management and the like, and the obtained actual demand proportion of each attribute combined commodity comprising new and existing commodities is used as a reference for class quantity configuration.
Regarding the former, regarding new sales prediction, using a machine learning algorithm in combination with the thought of sample transfer learning is a common method, but the problem is that the number of features available for extraction of new products is small, which can affect the model training effect and further affect the prediction accuracy; regarding the latter, regarding the problem of class management, since there are relevance and substitution among commodities, sales shares of some commodities in the current store cannot represent real demands of consumers, and therefore cannot be directly used as references for quantity configuration, and in the field of operational planning, only the measurement of substitution effect is considered in a commodity actual demand prediction model which is generally adopted, and promotion effect brought by relevance to commodity demands is rarely considered.
In view of this, the present invention is specifically proposed.
Disclosure of Invention
In order to solve the technical problems, the invention provides a multi-class combined demand prediction method and device based on substitution and association, which are used for establishing a demand prediction model based on the data analysis capability of machine learning and the planning capability of operation planning technology and obtaining the actual demand prediction result of commodities, assisting merchants to introduce new products and configuration decisions of the number of the commodities and enabling class management.
Specifically, the following technical scheme is adopted:
the multi-class joint demand prediction method based on substitution and association effects comprises the following steps:
based on the sales information of the currently on-sale commodity, determining the core attribute of the commodity, and converting the SKU-level commodity into an attribute-level commodity;
defining substitution and association actions among all attribute combination commodities in the purchasing behavior of consumers;
and incorporating parameters for measuring the replacement and the association into an operation planning model established based on the maximum likelihood idea, and solving the actual demand and the demand proportion of all attribute combined commodities comprising the existing commodity and the new commodity by using a meta heuristic algorithm.
In the multi-category joint demand prediction method based on substitution and association according to the present invention, the determining the core attribute of the commodity based on the sales information of the commodity currently being sold, and converting the SKU level commodity into the attribute level commodity includes:
extracting attribute characteristics of the currently on-sale commodity, including commodity specification category, and/or price category, and/or brand grade category, and/or brand place category, and/or function category;
and training a random forest regression model by taking the corresponding periodic sales volume of the currently sold commodity as a label, determining the core attribute of the commodity through marginal contribution degree ordering of the features based on the interpretable SHAP model, and converting the SKU-level commodity into the attribute-level commodity.
In an alternative embodiment of the present invention, in the multi-class joint demand prediction method based on substitution and association, the defining substitution among the combined commodities with each attribute in the purchasing behavior of the consumer includes:
based on the combination of different attributes of the current sold commodities, the substitution effect among the commodities of the same class is measured, and the substitution probability among the commodities of the corresponding attribute combination is determined as the product of the substitution probabilities among the attributes by referring to the thought of the joint analysis.
In an alternative embodiment of the present invention, in the multi-class joint demand prediction method based on substitution and association, the calculation process of the substitution probability includes:
the commodity set corresponding to a certain class is S, the favorite commodity of the user is i, when i is not in the class set S, the user can replace the similar commodity j of i in the class set S, and the replacement probability is pi ij The method comprises the steps of carrying out a first treatment on the surface of the Then rank j on A attributes respectively 1 ，j 2 ，...，j A J commodity substitutions are rated i on A attributes 1 ，i 2 ，...，i A The probability of the i commodity of (2) is
In the multi-class joint demand prediction method based on substitution and association according to the present invention, the defining the association between the combined commodities of each attribute in the purchasing behavior of the consumer includes:
determining association class and association probability based on association rules obtained by an Apriori algorithm, and assuming the association probability as a confidence coefficient result corresponding to the association rule mining;
and supposing that the association rule is used for mining the determined commodities with the first class corresponding to the m attribute combinations, and the second class corresponding to the commodities with the n attribute combinations, obtaining estimated values of mn association probabilities according to the record of purchasing the receipt by the user, and changing the commodity names in the receipt data into attribute level names during association rule mining to obtain the association probabilities corresponding to the commodities with the various attribute combinations across the classes.
As an alternative embodiment of the invention, the multi-class joint demand prediction method based on substitution and association comprises the following steps:
the attribute combination commodity sets of a pair of related classes are S and Z respectively, commodity p of each attribute combination in the set Z has a related effect on commodity j of each attribute combination in the set S, and the related probability is w jp The probability of correlation of commodity q not combined with each attribute in set Z to commodity j combined with each attribute in set S is w jp π qp 。
In the multi-category combined demand prediction method based on substitution and association according to the present invention, the step of incorporating parameters for measuring substitution and association into an operational planning model established based on maximum likelihood thinking, and the step of solving actual demand and demand proportion of each attribute combined commodity including existing commodity and new commodity by using meta heuristic algorithm includes:
the method comprises the steps of incorporating core attribute parameters, association probability, substitution probability and probability of favorite certain attribute combined commodities into an operation planning model established based on maximum likelihood ideas:
the attribute combined commodity sets of a pair of related classes are S and Z respectively, and the probability F of purchasing the attribute combined commodity j in the S set j (S) isProbability F of purchasing property combination commodity p in collection Z p (Z) the same theory, the objective function of the maximum likelihood estimation isWherein the method comprises the steps off j Probability, y of favorite commodity j P 、x j Store sales for commodity p, j respectively;
and solving the actual demand of each attribute combined commodity comprising the existing commodity and the new commodity by using a meta-heuristic algorithm.
As an alternative embodiment of the invention, the multi-class joint demand prediction method based on substitution and association comprises the following steps:
and determining the association probability among the associated categories based on a confidence result obtained by mining association rules by an Apriori algorithm, and establishing the association among a plurality of categories, thereby establishing a multi-category joint demand prediction model, wherein an objective function is the product of the purchasing probability of each category to the power, and the power items are the store sales of the combined commodity with the corresponding attribute of each category.
The invention also provides a multi-class joint demand prediction device based on substitution and association, which comprises:
the commodity attribute module is used for determining the core attribute of the commodity based on the sales information of the commodity currently being sold and converting the SKU-level commodity into an attribute-level commodity;
the substitution and association module is used for defining substitution and association actions among all attribute combination commodities in the purchasing behavior of the consumer;
and the commodity demand prediction module is used for incorporating parameters for measuring substitution and association into an operation planning model established based on a maximum likelihood idea, and solving the actual demand and demand proportion of all attribute combined commodities comprising the existing commodity and the new commodity by using a meta heuristic algorithm.
The invention also provides a computer storage medium which stores a computer executable program, and is characterized in that when the computer executable program is executed, the multi-class joint demand prediction method based on substitution and association is realized.
Compared with the prior art, the invention has the beneficial effects that:
the multi-category combined demand prediction method based on the substitution and association effect is suitable for predicting the actual demand (without being influenced by the substitution and association effect) and the new demand of the existing commodity in the store of the retail store, considers the substitution and association effect in the purchase behavior of the consumer, and is respectively as follows: selecting substitution when no favorite commodity of the user exists in the current class set; for related commodities which exist in the current store class set at the same time, the user has a certain probability of selecting the association; for associated items that are not concurrently present in the current set of store categories, the promoting effect of one party not in the set on the other party's initial demand is reflected in the product of the substitution probability and the association probability, i.e., the portion of the association promoting effect that is transferred to its substitution item. Therefore, the multi-class combined demand prediction method based on the substitution and association effect can obtain more accurate commodity actual demand prediction results by more completely defining the purchasing behavior of consumers in the model.
The invention provides a multi-class combined demand prediction method based on substitution and association, wherein a demand prediction model is expanded to be suitable for the situation of two/more associated classes, association probability estimated values obtained through association rule mining establish the association among all associated classes, the association is considered more completely, and the accuracy of commodity actual demand prediction is improved.
Therefore, the multi-category combined demand prediction method based on the substitution and association effect uses the historical sales data of the combined commodities with different attributes of each category of retail stores, and realizes multi-category combined actual demand prediction by defining the purchasing behavior of consumers including the substitution and association effect and establishing a demand prediction model applicable to a plurality of associated categories. Meanwhile, the demand proportion determined according to the actual demand of the existing commodity and the new commodity can be used for class planning guidance of retail stores of different scales.
Description of the drawings:
FIG. 1 is a flow chart of a multi-class joint demand prediction method based on surrogate and associative roles as disclosed in an embodiment of the present invention;
FIG. 2 is a flow chart of a multi-class joint demand prediction model establishment based on surrogate and associative roles as disclosed in an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. It will be apparent that the described embodiments are some, but not all, embodiments of the invention.
Thus, the following detailed description of the embodiments of the invention is not intended to limit the scope of the invention, as claimed, but is merely representative of some embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that, under the condition of no conflict, the embodiments of the present invention and the features and technical solutions in the embodiments may be combined with each other.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.
In the description of the present invention, it should be noted that, the terms "upper", "lower", and the like indicate an azimuth or a positional relationship based on the azimuth or the positional relationship shown in the drawings, or an azimuth or a positional relationship conventionally put in use of the inventive product, or an azimuth or a positional relationship conventionally understood by those skilled in the art, such terms are merely for convenience of describing the present invention and simplifying the description, and do not indicate or imply that the apparatus or element to be referred must have a specific azimuth, be constructed and operated in a specific azimuth, and thus should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and the like, are used merely to distinguish between descriptions and should not be construed as indicating or implying relative importance.
Referring to fig. 1, the multi-class joint demand prediction method based on substitution and association according to the embodiment includes:
screening and determining core properties of commodities: based on the sales information of the currently on-sale commodity, determining the core attribute of the commodity, and converting the SKU-level commodity into an attribute-level commodity;
defining substitution and association actions among all attribute combination commodities in the purchasing behavior of consumers;
and incorporating parameters for measuring the replacement and the association into an operation planning model established based on the maximum likelihood idea, and solving the actual demand and the demand proportion of all attribute combined commodities comprising the existing commodity and the new commodity by using a meta heuristic algorithm.
The multi-category combined demand prediction method based on the substitution and association actions is suitable for predicting the actual demand (not influenced by the substitution and association actions) of existing commodities and predicting new demand of retail stores, historical sales data of combined commodities with different attributes of each category of the retail stores is used, and the multi-category combined actual demand prediction is realized by defining consumer purchasing behaviors including the substitution and association actions, and establishing a demand prediction model suitable for a plurality of associated categories. Meanwhile, the demand proportion determined according to the actual demand of the existing commodity and the new commodity can be used for class planning guidance of retail stores of different scales.
According to the multi-class joint demand prediction method based on the substitution and association, an SHAP model is used for carrying out explanatory analysis on a random forest regression model when determining commodity core attributes, an Apr i or i algorithm is used when defining association among cross-class commodities, and a differential evolution algorithm is used when solving the model.
The random forest regression model is composed of a plurality of regression trees, and each decision tree in the forest is not associated with each other, and the final output of the model is jointly determined by each decision tree in the forest. The model introduces randomness in the training process of the decision tree, so that the model has excellent overfitting resistance and noise resistance.
The SHAP model is an interpretable machine learning model, can interpret the output of any machine learning model, and is named as Shap l ey add it i ve exp l anat i on, the SHAP builds an additive interpretation model under the heuristic of the cooperative game theory, all features are regarded as 'contributors', for each prediction sample, the model generates a predicted value, and SHAP va l ue is the value assigned to each feature in the sample, and can reflect the influence of the feature in each sample.
The apri or i algorithm is an association rule mining algorithm, and decomposes an association rule mining task into two main subtasks: subtask 1 is to generate frequent item sets and subtask 2 is to generate strong association rules. The support and the confidence are used for quantifying the frequent item set and the association rule respectively, and the Apr i or i algorithm is used for mining association classes with higher confidence in the strong association rule on the definition of the association effect in the embodiment.
The differential evolution algorithm is a heuristic search algorithm based on a population, the evolution process comprises mutation, hybridization and selection operations, the differential evolution algorithm mutation vector is generated by a father differential vector and is crossed with a father individual vector to generate a new individual vector, and the new individual vector is directly selected with the father individual, so that the risk of falling into a local optimal solution can be avoided with higher probability by the mutation operation compared with the random mutation operation of a genetic algorithm, namely, a certain distance between the original solutions is ensured by scaling, the exploration capability is ensured, and a better prediction result is facilitated by using the differential evolution algorithm to solve a demand prediction model.
Referring to fig. 2, in the multi-category joint demand prediction method based on substitution and association according to the present embodiment, determining a core attribute of a commodity based on sales information of a commodity currently being sold, and converting a SKU-level commodity into an attribute-level commodity includes:
extracting attribute characteristics of the currently on-sale commodity, including commodity specification categories (large, medium and small), price categories (high, medium and low), brand grade categories (brand network on-board/off-board), brand origin categories (yes/no local commodity) and/or functional categories;
and training a random forest regression model by taking the corresponding periodic sales quantity of the currently sold commodity as a label, determining the core attribute of the commodity through marginal contribution degree ordering of the features based on an interpretable SHAP model, and converting the SKU (stock level unit) level commodity into the attribute level commodity. For example, selecting core attributes as brand grade attribute, specification category and price category, then "25 g of chocolate flavor of a piglet petty sauce cup with a price per g of 0.12 yuan" can be named as "not in a bang" according to the specification and price distribution situation of biscuit type commodities.
According to the embodiment, SKU-level commodities are converted into attribute-level commodities, and based on the attribute-level commodities, the substitution probability among all attribute combination commodities of the same class and the association probability among all attribute combination commodities of the cross-class are determined.
Further, in the multi-class joint demand prediction method based on substitution and association according to this embodiment, the defining substitution among the combined commodities of each attribute in the purchasing behavior of the consumer includes:
based on the combination of different attributes of the current sold commodities, the substitution effect among the commodities of the same class is measured, and the substitution probability among the commodities of the corresponding attribute combination is determined as the product of the substitution probabilities among the attributes by referring to the thought of the joint analysis.
Specifically, in the multi-class joint demand prediction method based on substitution and association according to the embodiment, the calculation process of the substitution probability includes:
the commodity set corresponding to a certain class is S, the favorite commodity of the user is i, when i is not in the class set S, the user can replace the similar commodity j of i in the class set S, and the replacement probability is pi ij The method comprises the steps of carrying out a first treatment on the surface of the Then rank j on A attributes respectively 1 ，j 2 ，...，j A J commodity substitutions are rated i on A attributes 1 ，i 2 ，...，i A The probability of the i commodity of (2) is
As an optional implementation manner of the present embodiment, in the multi-class joint demand prediction method based on substitution and association in the present embodiment, the defining association between the combined commodities of each attribute in the purchasing behavior of the consumer includes:
determining association class and association probability based on association rules obtained by an Apriori algorithm, and assuming the association probability as a confidence coefficient result corresponding to the association rule mining;
and supposing that the association rule is used for mining the determined commodities with the first class corresponding to the m attribute combinations, and the second class corresponding to the commodities with the n attribute combinations, obtaining estimated values of mn association probabilities according to the record of purchasing the receipt by the user, and changing the commodity names in the receipt data into attribute level names during association rule mining to obtain the association probabilities corresponding to the commodities with the various attribute combinations across the classes.
Specifically, in the multi-class joint demand prediction method based on substitution and association in this embodiment, the calculation method of association probability includes:
the attribute combination commodity sets of a pair of related classes are S and Z respectively, commodity p of each attribute combination in the set Z has a related effect on commodity j of each attribute combination in the set S, and the related probability is w jp The probability of correlation of commodity q not combined with each attribute in set Z to commodity j combined with each attribute in set s is w jp π qp 。
According to the multi-category combined demand prediction method based on the substitution and association effect, parameters for measuring the substitution and association effect are brought into an operation planning model established based on a maximum likelihood idea, and a meta heuristic algorithm is used for solving actual demand and demand proportion of all attribute combined commodities including existing commodities and new commodities, wherein the method comprises the following steps:
and (3) incorporating the commodity core attribute parameters, the association probability, the substitution probability and the probability of favoring a certain attribute combined commodity into an operation planning model established based on the maximum likelihood idea. Assuming that past sales scenarios already contain consideration for substitution and association, i.e., the user's probability of purchasing any item in the current collection of items can be divided into three parts: because of the increased purchase probability of substitution, the purchase probability of association promotion, the purchase probability not affected by substitution and association, it is considered that data samples extracted periodically in a certain time range reflect the overall situation (Max problem) with the maximum probability.
The attribute combination commodity sets of a pair of related classes are S and Z respectively for purchaseProbability F of attribute combination commodity j in set S j (S) isProbability F of purchasing property combination commodity p in collection Z p (Z) the same theory, the objective function of the maximum likelihood estimation isWherein f j Probability, y of favorite commodity j P 、x j The store sales for commodity p, j, respectively.
And solving the actual demand of each attribute combined commodity comprising the existing commodity and the new commodity by using a meta-heuristic algorithm.
Preferably, solving the actual demand of each attribute combined commodity comprising the existing commodity and the new commodity by using a differential evolution algorithm;
optionally, the algorithm parameters that need to be searched are population size (NP), scaling factor (F), crossover probability (CR).
The multi-class joint demand prediction method based on substitution and association according to the embodiment includes:
and determining the association probability among the associated categories based on a confidence result obtained by mining association rules by an Apriori algorithm, and establishing the association among a plurality of categories, thereby establishing a multi-category joint demand prediction model, wherein an objective function is the product of the purchasing probability of each category to the power, and the power items are the store sales of the combined commodity with the corresponding attribute of each category.
The prediction result processing of the multi-class joint demand prediction method based on substitution and association in the embodiment: and determining respective demand proportions according to the actual demand pre-measurement of the new product and the existing product, and guiding the class planning problem of the retail store.
The method can obtain more accurate commodity actual demand prediction results by more completely defining the consumer purchasing behavior in a model, the demand prediction model is expanded to be suitable for the situation of two/more associated classes, the association probability estimated values obtained by association rule mining establish the association among all the associated classes, the association consideration is more complete, and the accuracy of commodity actual demand prediction is further improved.
According to the multi-class joint demand prediction method based on substitution and association, historical sales data of retail stores are used, and demand prediction is achieved through parameter definition, model establishment and solving. Identifying attributes having a key impact on sales of the commodity through an interpretable machine learning model; by more completely identifying the purchasing behavior of the consumer in the model, namely defining substitution and association, the model parameter definition is realized, the connection between different classes is established, and the accuracy of demand prediction by adopting the method is improved by the parameter definition process. The model is established by adopting the maximum likelihood estimation principle, the actual demand of the combined commodity with each attribute including the new commodity and the existing commodity is solved based on the differential evolution algorithm, and the demand proportion determined according to the actual demand of the existing commodity and the new commodity can be used for class planning guidance of retail stores with different scales.
The embodiment also provides a multi-class joint demand prediction device based on substitution and association, which comprises:
the commodity attribute module is used for determining the core attribute of the commodity based on the sales information of the commodity currently being sold and converting the SKU-level commodity into an attribute-level commodity;
the substitution and association module is used for defining substitution and association actions among all attribute combination commodities in the purchasing behavior of the consumer;
and the commodity demand prediction module is used for incorporating parameters for measuring substitution and association into an operation planning model established based on a maximum likelihood idea, and solving the actual demand and demand proportion of all attribute combined commodities comprising the existing commodity and the new commodity by using a meta heuristic algorithm.
The present embodiment also provides a computer storage medium storing a computer executable program which, when executed, implements the multi-class joint demand prediction method based on substitution and association actions as described above.
The computer storage medium of this embodiment may include a data signal propagated in baseband or as part of a carrier wave with readable program code embodied therein. 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 storage medium may also be any readable medium 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 storage 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.
The embodiment also provides an electronic device, which comprises a processor and a memory, wherein the memory is used for storing a computer executable program, and when the computer program is executed by the processor, the processor executes the multi-class joint demand prediction method based on substitution and association.
The electronic device is in the form of a general purpose computing device. The processor may be one or a plurality of processors and work cooperatively. The invention does not exclude that the distributed processing is performed, i.e. the processor may be distributed among different physical devices. The electronic device of the present invention is not limited to a single entity, but may be a sum of a plurality of entity devices.
The memory stores a computer executable program, typically machine readable code. The computer readable program may be executable by the processor to enable an electronic device to perform the method, or at least some of the steps of the method, of the present invention.
The memory includes volatile memory, such as Random Access Memory (RAM) and/or cache memory, and may be non-volatile memory, such as Read Only Memory (ROM).
It should be understood that elements or components not shown in the above examples may also be included in the electronic device of the present invention. For example, some electronic devices further include a display unit such as a display screen, and some electronic devices further include a man-machine interaction element such as a button, a keyboard, and the like. The electronic device may be considered as covered by the invention as long as the electronic device is capable of executing a computer readable program in a memory for carrying out the method or at least part of the steps of the method.
From the above description of embodiments, those skilled in the art will readily appreciate that the present invention may be implemented by hardware capable of executing a specific computer program, such as the system of the present invention, as well as electronic processing units, servers, clients, handsets, control units, processors, etc. included in the system. The invention may also be implemented by computer software executing the method of the invention, e.g. by control software executed by a microprocessor, an electronic control unit, a client, a server, etc. It should be noted, however, that the computer software for performing the method of the present invention is not limited to being executed by one or a specific hardware entity, but may also be implemented in a distributed manner by unspecified specific hardware. For computer software, the software product may be stored on a computer readable storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.), or may be stored distributed over a network, as long as it enables the electronic device to perform the method according to the invention.
The above embodiments are only for illustrating the present invention and not for limiting the technical solutions described in the present invention, and although the present invention has been described in detail in the present specification with reference to the above embodiments, the present invention is not limited to the above specific embodiments, and thus any modifications or equivalent substitutions are made to the present invention; all technical solutions and modifications thereof that do not depart from the spirit and scope of the invention are intended to be included in the scope of the appended claims.
Claims (10)
1. The multi-class joint demand prediction method based on substitution and association is characterized by comprising the following steps:
based on the sales information of the currently on-sale commodity, determining the core attribute of the commodity, and converting the SKU-level commodity into an attribute-level commodity;
defining substitution and association actions among all attribute combination commodities in the purchasing behavior of consumers;
and incorporating parameters for measuring the replacement and the association into an operation planning model established based on the maximum likelihood idea, and solving the actual demand and the demand proportion of all attribute combined commodities comprising the existing commodity and the new commodity by using a meta heuristic algorithm.
2. The multi-category joint demand prediction method based on substitution and association according to claim 1, wherein the determining the core attribute of the commodity based on the sales information of the commodity currently being sold, converting the SKU-level commodity into the attribute-level commodity comprises:
extracting attribute characteristics of the currently on-sale commodity, including commodity specification category, and/or price category, and/or brand grade category, and/or brand place category, and/or function category;
and training a random forest regression model by taking the corresponding periodic sales volume of the currently sold commodity as a label, determining the core attribute of the commodity through marginal contribution degree ordering of the features based on the interpretable SHAP model, and converting the SKU-level commodity into the attribute-level commodity.
3. The multi-category joint demand prediction method based on substitution and association according to claim 1, wherein the defining the substitution among the combined commodities of each attribute in the purchasing behavior of the consumer comprises:
based on the combination of different attributes of the current sold commodities, the substitution effect among the commodities of the same class is measured, and the substitution probability among the commodities of the corresponding attribute combination is determined as the product of the substitution probabilities among the attributes by referring to the thought of the joint analysis.
4. A multi-class joint demand prediction method based on substitution and correlation according to claim 3, wherein the calculation process of substitution probability comprises:
the commodity set corresponding to a certain class is S, the favorite commodity of the user is i, whenWhen i is not in the class set S, the user may replace the similar commodity j of i in the class set S with a substitution probability of pi ij The method comprises the steps of carrying out a first treatment on the surface of the Then rank j on A attributes respectively 1 ，j 2 ，...，j A J commodity substitutions are rated i on A attributes 1 ，i 2 ，...，i A The probability of the i commodity of (2) is
5. The multi-class joint demand prediction method based on substitution and association according to claim 3 or 4, wherein the defining the association between the combined commodities of each attribute in the purchasing behavior of the consumer comprises:
determining association class and association probability based on association rules obtained by an Apriori algorithm, and assuming the association probability as a confidence coefficient result corresponding to the association rule mining;
and supposing that the association rule is used for mining the determined commodities with the first class corresponding to the m attribute combinations, and the second class corresponding to the commodities with the n attribute combinations, obtaining estimated values of mn association probabilities according to the record of purchasing the receipt by the user, and changing the commodity names in the receipt data into attribute level names during association rule mining to obtain the association probabilities corresponding to the commodities with the various attribute combinations across the classes.
6. The multi-class joint demand prediction method based on substitution and correlation according to claim 5, comprising:
the attribute combination commodity sets of a pair of related classes are S and Z respectively, commodity p of each attribute combination in the set Z has a related effect on commodity j of each attribute combination in the set S, and the related probability is w jp The probability of association of commodity q not combined with each attribute in set Z to commodity J combined with each attribute in set S is w jp π qp 。
7. The multi-class joint demand prediction method based on substitution and correlation according to claim 5, wherein the step of incorporating the parameters for measuring substitution and correlation into the operation planning model established based on maximum likelihood thinking, and the step of solving the actual demand and the demand proportion of each attribute combined commodity including the existing commodity and the new commodity by using meta-heuristic algorithm comprises:
the method comprises the steps of incorporating core attribute parameters, association probability, substitution probability and probability of favorite certain attribute combined commodities into an operation planning model established based on maximum likelihood ideas:
the attribute combined commodity sets of a pair of related classes are S and Z respectively, and the probability F of purchasing the attribute combined commodity j in the S set j (S) isProbability F of purchasing property combination commodity p in collection Z p (Z) the same theory, the objective function of the maximum likelihood estimation isWherein f j Probability, y of favorite commodity j P 、x j Store sales for commodity p, j respectively;
and solving the actual demand of each attribute combined commodity comprising the existing commodity and the new commodity by using a meta-heuristic algorithm.
8. The multi-class joint demand prediction method based on substitution and correlation according to claim 7, comprising:
and determining the association probability among the associated categories based on a confidence result obtained by mining association rules by an Apriori algorithm, and establishing the association among a plurality of categories, thereby establishing a multi-category joint demand prediction model, wherein an objective function is the product of the purchasing probability of each category to the power, and the power items are the store sales of the combined commodity with the corresponding attribute of each category.
9. The multi-class joint demand prediction device based on substitution and association effects is characterized by comprising:
the commodity attribute module is used for determining the core attribute of the commodity based on the sales information of the commodity currently being sold and converting the SKU-level commodity into an attribute-level commodity;
the substitution and association module is used for defining substitution and association actions among all attribute combination commodities in the purchasing behavior of the consumer;
and the commodity demand prediction module is used for incorporating parameters for measuring substitution and association into an operation planning model established based on a maximum likelihood idea, and solving the actual demand and demand proportion of all attribute combined commodities comprising the existing commodity and the new commodity by using a meta heuristic algorithm.
10. A computer storage medium storing a computer executable program, wherein the computer executable program when executed implements the substitution and association based multi-class joint demand prediction method according to any one of claims 1-8.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2022116555550 | 2022-12-22 | ||
| CN202211655555.0A CN115809891A (en) | 2022-12-22 | 2022-12-22 | Multi-product combined demand prediction method and device based on substitution and correlation effects |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116579803A true CN116579803A (en) | 2023-08-11 |
Family
ID=85486659
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211655555.0A Pending CN115809891A (en) | 2022-12-22 | 2022-12-22 | Multi-product combined demand prediction method and device based on substitution and correlation effects |
| CN202310380958.7A Pending CN116579803A (en) | 2022-12-22 | 2023-04-11 | Multi-class joint demand prediction method and device based on substitution and association effects |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211655555.0A Pending CN115809891A (en) | 2022-12-22 | 2022-12-22 | Multi-product combined demand prediction method and device based on substitution and correlation effects |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN115809891A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117787867A (en) * | 2024-02-27 | 2024-03-29 | 山东财经大学 | Medicine inventory demand analysis method and system |
-
2022
- 2022-12-22 CN CN202211655555.0A patent/CN115809891A/en active Pending
-
2023
- 2023-04-11 CN CN202310380958.7A patent/CN116579803A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117787867A (en) * | 2024-02-27 | 2024-03-29 | 山东财经大学 | Medicine inventory demand analysis method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115809891A (en) | 2023-03-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9064212B2 (en) | Automatic event categorization for event ticket network systems | |
| US11227217B1 (en) | Entity transaction attribute determination method and apparatus | |
| EP4242955A1 (en) | User profile-based object recommendation method and device | |
| US20190080352A1 (en) | Segment Extension Based on Lookalike Selection | |
| CN111027895A (en) | Stock prediction and behavior data collection method, apparatus, device and medium for commodity | |
| Yan et al. | Implementation of a product-recommender system in an IoT-based smart shopping using fuzzy logic and apriori algorithm | |
| US20180012284A1 (en) | Information analysis apparatus, information analysis method, and non-transitory computer readable storage medium | |
| CN112365283B (en) | Coupon issuing method and device, terminal equipment and storage medium | |
| Saleh | Machine Learning Fundamentals: Use Python and scikit-learn to get up and running with the hottest developments in machine learning | |
| Kumar et al. | A computational framework for ranking prediction of cloud services under fuzzy environment | |
| CN116579803A (en) | Multi-class joint demand prediction method and device based on substitution and association effects | |
| CN113781149A (en) | Information recommendation method and device, computer-readable storage medium and electronic equipment | |
| US20230076083A1 (en) | Methods and apparatus for generating training data to train machine learning based models | |
| WO2022156589A1 (en) | Method and device for determining live broadcast click rate | |
| US11494593B2 (en) | Methods and apparatus for machine learning model hyperparameter optimization | |
| CN111461757A (en) | Information processing method and device, computer storage medium and electronic equipment | |
| CN111026973A (en) | Commodity interest degree prediction method and device and electronic equipment | |
| US20230076149A1 (en) | Methods and apparatus for data imputation of a sparse time series data set | |
| CN112632275B (en) | Crowd clustering data processing method, device and equipment based on personal text information | |
| CN113792952A (en) | Method and apparatus for generating a model | |
| CN112085093A (en) | Training method and device of collaborative filtering model, readable medium and system | |
| Sudarsan et al. | E-commerce Website with Image Search and Price Prediction | |
| CN113159877A (en) | Data processing method, device, system and computer readable storage medium | |
| JP6809148B2 (en) | Program and combination extraction system | |
| CN111275176A (en) | Distributed computing method and distributed computing system |
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 |