CN117436936B - Sales prediction and BOM finished product processing system and method - Google Patents
Sales prediction and BOM finished product processing system and method Download PDFInfo
- Publication number
- CN117436936B CN117436936B CN202311748505.1A CN202311748505A CN117436936B CN 117436936 B CN117436936 B CN 117436936B CN 202311748505 A CN202311748505 A CN 202311748505A CN 117436936 B CN117436936 B CN 117436936B
- Authority
- CN
- China
- Prior art keywords
- sales
- feature
- target
- feature vector
- data
- 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.)
- Active
Links
- 238000012545 processing Methods 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 title claims abstract description 16
- 239000013598 vector Substances 0.000 claims abstract description 182
- 238000012549 training Methods 0.000 claims abstract description 80
- 238000004458 analytical method Methods 0.000 claims abstract description 28
- 230000035772 mutation Effects 0.000 claims description 43
- 238000004364 calculation method Methods 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 4
- 230000003203 everyday effect Effects 0.000 claims description 4
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000013139 quantization Methods 0.000 description 3
- 238000012937 correction Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000001932 seasonal effect Effects 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
- 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; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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
- G06N3/08—Learning methods
-
- 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
- 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; 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- 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)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- General Engineering & Computer Science (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biophysics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Human Resources & Organizations (AREA)
- Software Systems (AREA)
- Tourism & Hospitality (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Manufacturing & Machinery (AREA)
- Primary Health Care (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a sales volume prediction and BOM finished product processing system and method, and relates to the technical field of data processing. The method comprises the following steps: acquiring historical sales data of a target product, extracting a plurality of feature vectors, constructing a first training data set and a second training data set, and training to obtain a sales prediction model and a feature difference analysis model; obtaining a first target feature vector in a first preset time range and a second target feature vector in a second preset time range, generating first predicted sales data through a sales prediction model, generating a target sales difference factor through a feature difference analysis model, and correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data. The invention improves the efficiency of warehouse operation and is convenient for managing the processing of target products.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a sales volume prediction and BOM finished product processing system and method.
Background
In the current e-commerce logistics operation, warehouse operation is performed based on real-time ordering quantity of user ordering, warehouse sorting main pipes generally predict according to personal experience and historical sales quantity conditions to improve warehouse operation efficiency, more experience data are needed in the mode, prediction difficulty is high when the data quantity is large, prediction accuracy is low, and accordingly sorting of commodity parts and subsequent finished product processing efficiency are low.
Disclosure of Invention
Aiming at the problems, the application provides a sales prediction and BOM finished product processing system and method based on sales prediction, which are used for predicting sales of commodities through a deep learning model to generate a BOM table of target products, so that the processing management of the target products is facilitated, and the efficiency of warehouse operation is improved.
As one aspect of the present application, there is provided a sales prediction and BOM based finishing system comprising:
The data acquisition module is used for acquiring historical sales data of the target product, wherein the historical sales data comprises sales data and influence characteristic data of the target product every day in a preset historical event section;
The feature extraction module is used for extracting a plurality of feature vectors from the historical sales data based on the standard vectors;
The sample construction module is used for constructing a first training data set and a second training data set based on a plurality of feature vectors, each group of sample data in the first training data set comprises a single feature vector and sales volume data corresponding to the single feature vector, and each group of sample data in the second training data set comprises an influence feature difference factor and sales volume difference factor;
The model training module is used for training the sales quantity prediction model through the first training data set and training the characteristic difference analysis model through the second training data set;
The sales predicting module is used for obtaining a first target feature vector in a first preset time range and a second target feature vector in a second preset time range, processing the first target feature vector through the sales predicting model to obtain first predicted sales data in the first preset time range, calculating to obtain a target influence feature difference factor based on the first target feature vector and the second target feature vector, processing the target influence feature difference factor through the feature difference analyzing model to obtain a target sales difference factor, and correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data;
and the processing management module is used for generating a BOM table of the target product based on the second predicted sales volume data and managing the processing of the target product based on the BOM table.
Further, the constructing a first training data set and a second training data set based on the plurality of feature vectors includes:
For a plurality of feature vectors, sorting the feature vectors based on time sequence, constructing a sales volume curve graph based on the sorted feature vectors, and determining a plurality of first mutation feature vectors based on the sales volume curve graph;
Screening out a plurality of the first abrupt change feature vectors, and constructing the first training data set based on the remaining plurality of the feature vectors;
Determining a second mutation feature vector corresponding to each first mutation feature vector, calculating a plurality of influence feature difference factors based on each pair of associated first mutation feature vectors and second mutation feature vectors, calculating a plurality of sales difference factors based on sales data corresponding to each pair of associated first mutation feature vectors and second mutation feature vectors, and constructing a second training data set based on the plurality of influence feature difference factors and the plurality of sales difference factors.
Further, the determining a plurality of first abrupt feature vectors based on the sales volume graph includes:
For any two adjacent feature vectors And/>If the feature vector/>And the feature vectorCorresponding sales data/>And/>Satisfy the condition |/>If the I is larger than the preset difference threshold value, thenThe corresponding feature vector is the first abrupt feature vector;
The determining the second mutation feature vector corresponding to each first mutation feature vector comprises:
For any two adjacent feature vectors And/>If the feature vector/>And the feature vectorCorresponding sales data/>And/>Satisfy the condition |/>And if the I is larger than the preset difference threshold, the feature vector/>And the feature vector/>In/>The corresponding feature vector is the second abrupt feature vector.
Further, the calculating of the influence characteristic difference factor includes:
For any two mutation feature vectors, taking the similarity of the two mutation feature vectors as the influence feature difference factor, wherein the similarity of the two mutation feature vectors is one of cosine similarity, euclidean distance and Pearson correlation coefficient;
the calculation of the sales volume difference factor comprises the following steps:
for any associated abrupt feature vector And/>Taking the mutation feature vector/>And the mutation feature vector/>Corresponding sales data/>And/>The corresponding growth ratio is the sales difference factor.
Further, the correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data includes:
The calculation formula of the second predicted sales data is as follows:
；
In the method, in the process of the invention, For the second predicted sales data,/>For the first predicted sales data,/>Is the target sales volume difference factor.
Further, the sales volume prediction model is constructed based on a deep neural network, and the characteristic difference analysis model is constructed based on a convolutional neural network. For the sales volume prediction model, taking a plurality of feature vectors as input of the sales volume prediction model, taking sales volume data corresponding to each feature vector as a training target, and training to obtain the sales volume prediction model; and for the characteristic difference analysis model, taking a plurality of influence characteristic difference factors as input of the characteristic difference analysis model, and taking the sales volume difference factor corresponding to each influence characteristic difference factor as a training target, and training to obtain the characteristic difference analysis model.
As another aspect of the present application, there is provided a sales prediction and BOM product processing method based on the system implementation described above, including:
acquiring historical sales data of a target product, and extracting a plurality of feature vectors from the historical sales data based on standard vectors;
Constructing a first training data set and a second training data set based on a plurality of the feature vectors, training a sales prediction model through the first training data set, and training a feature difference analysis model through the second training data set;
Acquiring a first target feature vector in a first preset time range and a second target feature vector in a second preset time range, processing the first target feature vector through the sales prediction model to obtain first predicted sales data in the first preset time range, calculating a target influence feature difference factor based on the first target feature vector and the second target feature vector, processing the target influence feature difference factor through the feature difference analysis model to obtain a target sales difference factor, and correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data;
And generating a BOM table of the target product based on the second predicted sales data, and managing processing of the target product based on the BOM table.
The invention has the following advantages:
According to the method, a first training data set and a second training data set are constructed based on historical sales data, a sales prediction model and a characteristic difference analysis model are obtained through training, sales prediction of a target product is carried out through the sales prediction model, correction factors are determined through the characteristic difference analysis model, prediction output by the sales prediction model is corrected, a more accurate prediction result is obtained, processing of the target product is convenient to manage, and warehouse operation efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to the structures shown in these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a sales prediction and BOM product processing system according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, some embodiments of the present application will be described in further detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application. However, those of ordinary skill in the art will understand that in various embodiments of the present application, numerous technical details have been set forth in order to provide a better understanding of the present application. The claimed application may be practiced without these specific details and with various changes and modifications based on the following embodiments.
Referring to fig. 1, in an embodiment of the present invention, a sales prediction and BOM product processing system is provided, including:
The data acquisition module is used for acquiring historical sales data of the target product;
The historical sales data may include sales data and influence feature data of a target product every day in a preset historical event section, for example, sales data and influence feature data of a certain furniture every day in one month of history, where the influence feature data is specifically some factors that affect sales of the target product, for example, taking a certain assembled rocking chair as an example, a price of the product, a selling season of the product, an ambient temperature, a sales promotion activity of the product, a date type (such as holidays, workdays, holidays, etc.), and the like, and the influence feature of each type of product is different according to differences of attributes of the product, so that the type of the product is not limited in the implementation.
The feature extraction module is used for extracting a plurality of feature vectors from the historical sales data based on the standard vectors;
The feature vector includes quantization parameters of each influencing feature, for example, the content of the feature vector includes price change parameters, seasonal change parameters, discount change parameters, temperature change parameters, and the like, and the standard price can be determined corresponding to the standard vector, for example, for the price change parameters, the price change parameters in the feature vector on a certain day are determined according to the difference value between each actual price and the standard price, for example, the price reduction ratio, the price increase ratio, and the like, and the standard vector is established by technical means well known to those skilled in the art, and is not specifically limited herein.
A sample construction module for constructing a first training data set and a second training data set based on the plurality of feature vectors;
Each group of sample data in the first training data set comprises a single feature vector and sales volume data corresponding to the single feature vector, and each group of sample data in the second training data set comprises an influence feature difference factor and a sales volume difference factor, specifically, the influence feature difference factor is a difference quantization value between two feature vectors, and the sales volume difference factor is a change degree quantization value between sales volume data corresponding to each of the two feature vectors, such as an increase proportion, a decrease proportion and the like.
The model training module is used for training the sales prediction model through the first training data set and training the characteristic difference analysis model through the second training data set;
The feature difference analysis model is constructed based on a convolutional neural network, a plurality of feature vectors are used as input of the sales prediction model for the sales prediction model, sales data corresponding to each feature vector is used as a training target, and the sales prediction model is obtained through training and is used for predicting sales in a certain future time period by analyzing the feature vectors in the specific time period; and for the characteristic difference analysis model, taking a plurality of influence characteristic difference factors as the input of the characteristic difference analysis model, taking sales volume difference factors corresponding to each influence characteristic difference factor as training targets, and training to obtain the characteristic difference analysis model.
The sales predicting module is used for obtaining a first target feature vector in a first preset time range and a second target feature vector in a second preset time range, processing the first target feature vector through the sales predicting model to obtain first predicted sales data in the first preset time range, calculating a target influence feature difference factor based on the first target feature vector and the second target feature vector, processing the target influence feature difference factor through the feature difference analyzing model to obtain a target sales difference factor, and correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data;
The method comprises the steps of obtaining a first target feature vector of a target product tomorrow and a second target feature vector of the target product tomorrow, analyzing and processing the first target feature vector of the tomorrow through a sales prediction model to obtain first predicted sales data of the target product tomorrow, calculating to obtain a target influence feature difference factor through the first target feature vector of the tomorrow and the second target feature vector of the tomorrow, processing the target influence feature difference factor through a feature difference analysis model to obtain a target sales difference factor, and correcting the first predicted sales data of the target product tomorrow based on the target sales difference factor to obtain second predicted sales data of the target product tomorrow;
The processing management module is used for generating a BOM (material information table) of the target product based on the second predicted sales data, and managing the processing of the target product based on the BOM;
Specifically, after the second predicted sales data is determined, the required relevant material information can be determined based on the processing steps of the target product, so that a BOM table of the target product is obtained, the processing management of the target product is realized, and the efficiency of warehouse operation is improved.
In an alternative embodiment, for the sample construction module, constructing the first training data set and the second training data set based on the plurality of feature vectors specifically includes:
for a plurality of feature vectors, sorting the plurality of feature vectors based on a time sequence, constructing a sales volume graph based on the sorted plurality of feature vectors, and determining a plurality of first abrupt feature vectors based on the sales volume graph;
For example, for feature vectors of each day in the coming half year, the feature vectors are ordered according to the time development sequence, a coordinate system is constructed by taking the time sequence as a horizontal axis and sales data as a vertical axis, so that a sales curve graph of the sales changing along with time is obtained, and a plurality of first mutation feature vectors are determined according to a plurality of peaks of the sales curve graph.
Exemplary, for any two adjacent feature vectorsAnd/>If the feature vector/>And feature vectorCorresponding sales data/>And/>Satisfy the condition |/>If the I is larger than the preset difference threshold value, thenThe corresponding feature vector is the first abrupt feature vector.
Screening out a plurality of first abrupt change feature vectors, constructing a first training data set based on the rest of the plurality of feature vectors, specifically, determining sales data of each feature vector, associating each feature vector with corresponding sales data to obtain a plurality of groups of first sample data, and constructing the plurality of groups of first sample data to obtain the first training data set.
Determining a second mutation feature vector corresponding to each first mutation feature vector, and calculating a plurality of influence feature difference factors based on each pair of associated first mutation feature vectors and second mutation feature vectors;
Wherein for any two adjacent feature vectors And/>If the feature vector/>And feature vector/>Corresponding sales data/>And/>Satisfy the condition |/>If the I is larger than the preset difference threshold, the feature vector/>And feature vector/>In/>And the corresponding feature vector is the second mutation feature vector, and a plurality of influence feature difference factors are obtained through calculation according to a similarity calculation formula.
And calculating a plurality of sales volume difference factors based on sales volume data corresponding to each pair of associated first mutation feature vectors and second mutation feature vectors, and constructing a second training data set based on the plurality of influence feature difference factors and the plurality of sales volume difference factors.
Specifically, each influence characteristic difference factor is associated with a corresponding sales volume difference factor, a plurality of groups of second sample data are constructed, and a second training data set is constructed based on the plurality of groups of second sample data.
In an alternative embodiment, the calculation of the influencing feature difference factor comprises:
Taking any two mutation feature vectors as an example, taking the similarity of the two mutation feature vectors as an influence feature difference factor;
The similarity of the two mutation feature vectors may be one of cosine similarity, euclidean distance and pearson correlation coefficient, and in this embodiment, the cosine similarity of the two mutation feature vectors is calculated as an influence feature difference factor corresponding to the two mutation feature vectors by taking the cosine similarity as an example.
In an alternative embodiment, the calculation of the sales variance factor includes:
for any associated abrupt feature vector And/>Take mutation feature vector/>And mutation feature vectorCorresponding sales data/>And/>The corresponding growth ratio is a sales difference factor.
It is worth noting that sales dataAnd/>The corresponding growth ratio may be/>When/>Greater thanThe corresponding growth proportion is positive, indicating a growth, and otherwise negative, indicating a negative growth.
In an alternative embodiment, for the sales prediction module, correcting the first predicted sales data based on the target sales difference factor to obtain the second predicted sales data specifically includes:
the calculation formula of the second predicted sales data is as follows:
；
In the method, in the process of the invention, For the second predicted sales data,/>For the first predicted sales data,/>And calculating the second predicted sales data by the formula as a target sales difference factor.
According to the method, a first training data set and a second training data set are constructed based on historical sales data, a sales prediction model and a characteristic difference analysis model are obtained through training, sales prediction of a target product is carried out through the sales prediction model, correction factors are determined through the characteristic difference analysis model, prediction output by the sales prediction model is corrected, a more accurate prediction result is obtained, processing of the target product is convenient to manage, and warehouse operation efficiency is improved.
On the basis of the sales volume prediction and BOM finished product processing system provided in the embodiment of the invention, the embodiment of the invention also provides a sales volume prediction and BOM finished product processing method, which specifically comprises the following steps:
S1, acquiring historical sales data of a target product, and extracting a plurality of feature vectors from the historical sales data based on standard vectors;
s2, constructing a first training data set and a second training data set based on a plurality of feature vectors, training a sales prediction model through the first training data set, and training a feature difference analysis model through the second training data set;
s3, obtaining a first target feature vector in a first preset time range and a second target feature vector in a second preset time range, processing the first target feature vector through a sales prediction model to obtain first predicted sales data in the first preset time range, calculating a target influence feature difference factor based on the first target feature vector and the second target feature vector, processing the target influence feature difference factor through a feature difference analysis model to obtain a target sales difference factor, and correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data;
S4, generating a BOM table of the target product based on the second predicted sales data, and managing processing of the target product based on the BOM table.
It will be understood that modifications and variations will be apparent to those skilled in the art from the foregoing description, and it is intended that all such modifications and variations be included within the scope of the following claims. Parts of the specification not described in detail belong to the prior art known to those skilled in the art.
Claims (5)
1. A sales prediction and BOM product processing system comprising:
The data acquisition module is used for acquiring historical sales data of the target product, wherein the historical sales data comprises sales data and influence characteristic data of the target product every day in a preset historical event section;
The feature extraction module is used for extracting a plurality of feature vectors from the historical sales data based on the standard vectors;
The sample construction module is used for constructing a first training data set and a second training data set based on a plurality of feature vectors, each group of sample data in the first training data set comprises a single feature vector and sales volume data corresponding to the single feature vector, and each group of sample data in the second training data set comprises an influence feature difference factor and sales volume difference factor;
The model training module is used for training the sales quantity prediction model through the first training data set and training the characteristic difference analysis model through the second training data set;
The sales predicting module is used for obtaining a first target feature vector in a first preset time range and a second target feature vector in a second preset time range, processing the first target feature vector through the sales predicting model to obtain first predicted sales data in the first preset time range, calculating to obtain a target influence feature difference factor based on the first target feature vector and the second target feature vector, processing the target influence feature difference factor through the feature difference analyzing model to obtain a target sales difference factor, and correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data;
the processing management module is used for generating a BOM table of the target product based on the second predicted sales volume data and managing the processing of the target product based on the BOM table;
the constructing a first training data set and a second training data set based on the plurality of feature vectors includes:
For a plurality of feature vectors, sorting the feature vectors based on time sequence, constructing a sales volume curve graph based on the sorted feature vectors, and determining a plurality of first mutation feature vectors based on the sales volume curve graph;
Screening out a plurality of the first abrupt change feature vectors, and constructing the first training data set based on the remaining plurality of the feature vectors;
Determining a second mutation feature vector corresponding to each first mutation feature vector, calculating a plurality of influence feature difference factors based on each pair of associated first mutation feature vectors and second mutation feature vectors, calculating a plurality of sales difference factors based on sales data corresponding to each pair of associated first mutation feature vectors and second mutation feature vectors, and constructing a second training data set based on the plurality of influence feature difference factors and the plurality of sales difference factors;
The determining a plurality of first abrupt feature vectors based on the sales volume graph comprises:
For any two adjacent feature vectors And/>If the feature vector/>And the feature vector/>Corresponding sales data/>And/>Satisfy the condition |/>If I is greater than the preset difference threshold, then/>The corresponding feature vector is the first abrupt feature vector;
The determining the second mutation feature vector corresponding to each first mutation feature vector comprises:
For any two adjacent feature vectors And/>If the feature vector/>And the feature vector/>Corresponding sales data/>And/>Satisfy the condition |/>And if the I is larger than the preset difference threshold, the feature vector/>And the feature vector/>In/>The corresponding feature vector is the second abrupt feature vector.
2. The system of claim 1, wherein the calculation of the influencing feature difference factor comprises:
For any two mutation feature vectors, taking the similarity of the two mutation feature vectors as the influence feature difference factor, wherein the similarity of the two mutation feature vectors is one of cosine similarity, euclidean distance and Pearson correlation coefficient;
the calculation of the sales volume difference factor comprises the following steps:
for any associated abrupt feature vector And/>Taking the mutation feature vector/>And the mutation feature vector/>Corresponding sales data/>And/>The corresponding growth ratio is the sales difference factor.
3. The system of claim 2, wherein the correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data comprises:
The calculation formula of the second predicted sales data is as follows:
；
In the method, in the process of the invention, For the second predicted sales data,/>For the first predicted sales data,/>Is the target sales volume difference factor.
4. The system of claim 3, wherein the sales prediction model is constructed based on a deep neural network and the feature difference analysis model is constructed based on a convolutional neural network;
For the sales volume prediction model, taking a plurality of feature vectors as input of the sales volume prediction model, taking sales volume data corresponding to each feature vector as a training target, and training to obtain the sales volume prediction model; and for the characteristic difference analysis model, taking a plurality of influence characteristic difference factors as input of the characteristic difference analysis model, and taking the sales volume difference factor corresponding to each influence characteristic difference factor as a training target, and training to obtain the characteristic difference analysis model.
5. A method of finished product processing based on sales prediction and BOM, implemented based on the system of any one of claims 1-4, comprising:
acquiring historical sales data of a target product, and extracting a plurality of feature vectors from the historical sales data based on standard vectors;
Constructing a first training data set and a second training data set based on a plurality of the feature vectors, training a sales prediction model through the first training data set, and training a feature difference analysis model through the second training data set;
Acquiring a first target feature vector in a first preset time range and a second target feature vector in a second preset time range, processing the first target feature vector through the sales prediction model to obtain first predicted sales data in the first preset time range, calculating a target influence feature difference factor based on the first target feature vector and the second target feature vector, processing the target influence feature difference factor through the feature difference analysis model to obtain a target sales difference factor, and correcting the first predicted sales data based on the target sales difference factor to obtain second predicted sales data;
And generating a BOM table of the target product based on the second predicted sales data, and managing processing of the target product based on the BOM table.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311748505.1A CN117436936B (en) | 2023-12-19 | 2023-12-19 | Sales prediction and BOM finished product processing system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311748505.1A CN117436936B (en) | 2023-12-19 | 2023-12-19 | Sales prediction and BOM finished product processing system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117436936A CN117436936A (en) | 2024-01-23 |
| CN117436936B true CN117436936B (en) | 2024-05-17 |
Family
ID=89546498
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311748505.1A Active CN117436936B (en) | 2023-12-19 | 2023-12-19 | Sales prediction and BOM finished product processing system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117436936B (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017100278A1 (en) * | 2015-12-09 | 2017-06-15 | Wal-Mart Stores, Inc. | Systems and methods of utilizing multiple forecast models in forecasting customer demands for products at retail facilities |
| CN111538955A (en) * | 2020-04-17 | 2020-08-14 | 北京小米松果电子有限公司 | Goods sales prediction method, device and storage medium |
| CN111724211A (en) * | 2020-06-30 | 2020-09-29 | 名创优品（横琴）企业管理有限公司 | Offline store commodity sales prediction method, device and equipment |
| CN111815349A (en) * | 2020-05-28 | 2020-10-23 | 杭州览众数据科技有限公司 | Clothing sales volume trend prediction method based on historical similar products |
| CN113723985A (en) * | 2021-03-04 | 2021-11-30 | 京东城市(北京)数字科技有限公司 | Training method and device for sales prediction model, electronic equipment and storage medium |
| CN114022221A (en) * | 2021-11-25 | 2022-02-08 | 佛山欧神诺云商科技有限公司 | Sales prediction method, acquisition method and device of model thereof, and electronic equipment |
| CN114511273A (en) * | 2022-03-16 | 2022-05-17 | 湖南兴盛优选电子商务有限公司 | Retail warehousing management system and method |
| CN114757700A (en) * | 2022-04-12 | 2022-07-15 | 北京京东尚科信息技术有限公司 | Article sales prediction model training method, article sales prediction method and apparatus |
| WO2023016173A1 (en) * | 2021-08-10 | 2023-02-16 | 北京沃东天骏信息技术有限公司 | Inventory adjustment method and apparatus, electronic device, and computer readable medium |
| CN115907842A (en) * | 2022-10-24 | 2023-04-04 | 荣耀终端有限公司 | Product sales prediction method independent of self historical sales data and electronic equipment |
| CN115907839A (en) * | 2022-09-26 | 2023-04-04 | 浪潮通用软件有限公司 | Method, system, device and storage medium for predicting sales |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140278778A1 (en) * | 2013-03-15 | 2014-09-18 | Rangespan Limited | Method, apparatus, and computer-readable medium for predicting sales volume |
-
2023
- 2023-12-19 CN CN202311748505.1A patent/CN117436936B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017100278A1 (en) * | 2015-12-09 | 2017-06-15 | Wal-Mart Stores, Inc. | Systems and methods of utilizing multiple forecast models in forecasting customer demands for products at retail facilities |
| CN111538955A (en) * | 2020-04-17 | 2020-08-14 | 北京小米松果电子有限公司 | Goods sales prediction method, device and storage medium |
| CN111815349A (en) * | 2020-05-28 | 2020-10-23 | 杭州览众数据科技有限公司 | Clothing sales volume trend prediction method based on historical similar products |
| CN111724211A (en) * | 2020-06-30 | 2020-09-29 | 名创优品（横琴）企业管理有限公司 | Offline store commodity sales prediction method, device and equipment |
| CN113723985A (en) * | 2021-03-04 | 2021-11-30 | 京东城市(北京)数字科技有限公司 | Training method and device for sales prediction model, electronic equipment and storage medium |
| WO2023016173A1 (en) * | 2021-08-10 | 2023-02-16 | 北京沃东天骏信息技术有限公司 | Inventory adjustment method and apparatus, electronic device, and computer readable medium |
| CN114022221A (en) * | 2021-11-25 | 2022-02-08 | 佛山欧神诺云商科技有限公司 | Sales prediction method, acquisition method and device of model thereof, and electronic equipment |
| CN114511273A (en) * | 2022-03-16 | 2022-05-17 | 湖南兴盛优选电子商务有限公司 | Retail warehousing management system and method |
| CN114757700A (en) * | 2022-04-12 | 2022-07-15 | 北京京东尚科信息技术有限公司 | Article sales prediction model training method, article sales prediction method and apparatus |
| CN115907839A (en) * | 2022-09-26 | 2023-04-04 | 浪潮通用软件有限公司 | Method, system, device and storage medium for predicting sales |
| CN115907842A (en) * | 2022-10-24 | 2023-04-04 | 荣耀终端有限公司 | Product sales prediction method independent of self historical sales data and electronic equipment |
Non-Patent Citations (1)
| Title |
|---|
| 短生命周期产品的销量预测模型研究;赵学斌;李大学;谢名亮;;计算机工程与设计;20100616(11);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117436936A (en) | 2024-01-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110162693B (en) | Information recommendation method and server | |
| Le Ray et al. | Online adaptive clustering algorithm for load profiling | |
| Aouad et al. | Market segmentation trees | |
| CN105787061A (en) | Information pushing method | |
| Dai et al. | A clustering-based sales forecasting scheme using support vector regression for computer server | |
| CN116431931B (en) | Real-time incremental data statistical analysis method | |
| CN110555578A (en) | sales prediction method and device | |
| CN116385048B (en) | Intelligent marketing method and system for agricultural products | |
| CN112488863A (en) | Dangerous seed recommendation method and related equipment in user cold start scene | |
| CN116503142B (en) | Partner intelligent marketing scheduling data processing system | |
| CN111882363A (en) | Sales prediction method, system and terminal | |
| CN116304374B (en) | Customer matching method and system based on package data | |
| CN117436936B (en) | Sales prediction and BOM finished product processing system and method | |
| CN115841345B (en) | Cross-border big data intelligent analysis method, system and storage medium | |
| CN110599281A (en) | Method and device for determining target shop | |
| CN112330030A (en) | System and method for predicting requirements of expansion materials | |
| CN116703533A (en) | Business management data optimized storage analysis method | |
| CN115936184B (en) | Load prediction matching method suitable for multi-user types | |
| CN110766488B (en) | Method and device for automatically determining theme scenes | |
| CN114756758B (en) | Hybrid recommendation method and system | |
| Wang et al. | Using incremental clustering technique in collaborative filtering data update | |
| EP3493082A1 (en) | A method of exploring databases of time-stamped data in order to discover dependencies between the data and predict future trends | |
| CN110956528B (en) | Recommendation method and system for e-commerce platform | |
| CN111667290B (en) | Business display method and device and computer readable storage medium | |
| Sa | Collaborative filtering recommendation algorithm based on cloud model clustering of multi-indicators item evaluation |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |