CN118395135B - Data preprocessing system and method for supply chain planning system - Google Patents
Data preprocessing system and method for supply chain planning system Download PDFInfo
- Publication number
- CN118395135B CN118395135B CN202410866238.6A CN202410866238A CN118395135B CN 118395135 B CN118395135 B CN 118395135B CN 202410866238 A CN202410866238 A CN 202410866238A CN 118395135 B CN118395135 B CN 118395135B
- Authority
- CN
- China
- Prior art keywords
- data
- value data
- feature
- sequence
- feature value
- 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
- 238000013439 planning Methods 0.000 title claims abstract description 82
- 238000007781 pre-processing Methods 0.000 title claims abstract description 63
- 238000000034 method Methods 0.000 title claims abstract description 51
- 239000013598 vector Substances 0.000 claims abstract description 211
- 238000009826 distribution Methods 0.000 claims abstract description 80
- 238000005259 measurement Methods 0.000 claims abstract description 63
- 230000002159 abnormal effect Effects 0.000 claims abstract description 32
- 238000004458 analytical method Methods 0.000 claims abstract description 27
- 238000012549 training Methods 0.000 claims description 101
- 230000006870 function Effects 0.000 claims description 14
- 238000013135 deep learning Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 description 9
- 239000000463 material Substances 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 238000012384 transportation and delivery Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000001303 quality assessment method Methods 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000013068 supply chain management Methods 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a data preprocessing system and a method thereof for a supply chain planning system, wherein the system acquires a set of characteristic value data to be processed; carrying out association feature analysis based on a subset on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data; taking a certain inter-eigenvalue data associated eigenvector in the sequence of the inter-eigenvalue data associated eigenvector as a query eigenvector, and calculating a data distribution difference semantic measurement coefficient between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of the inter-eigenvalue data associated eigenvector; and determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the data distribution difference semantic measurement coefficient. In this way, the supply chain planning system may be aided in better understanding and managing the data, thereby improving the accuracy and efficiency of the decisions.
Description
Technical Field
The invention relates to the technical field of intelligent data preprocessing, in particular to a data preprocessing system and a method for a supply chain planning system.
Background
A supply chain planning system is a software system for managing various resources and activities in a supply chain network that can help businesses improve efficiency, reduce costs, increase profits, and meet customer needs. One of the core functions of the supply chain planning system is to generate optimal production, procurement, inventory and transportation plans based on predicted and actual demand. To achieve this, the supply chain planning system needs to process a large amount of data, including main data and business data. Master data refers to data describing entities and relationships in a supply chain network, such as materials, factories, warehouses, suppliers, customers, etc. Business data refers to data describing the operation in a supply chain network, such as orders, inventory, production, procurement, transportation, etc. These data are input to the supply chain planning system and are also the basis for the output plan. Therefore, the quality of the data is critical to the performance and effectiveness of the supply chain planning system.
However, in practical applications, supply chain planning systems face challenges of data quality issues. Due to the variety of data sources, complexity of data collection and transmission, difficulty in data updating and maintenance, etc., the data in the supply chain planning system may be out of specification, inconsistent, incomplete, inaccurate, etc. These problem data can affect the proper operation of the supply chain planning system, leading to erroneous or inefficient planning results, and thus, the operational efficiency and customer satisfaction of the enterprise. Accordingly, a data preprocessing system and method for a supply chain planning system is desired.
Disclosure of Invention
The embodiment of the invention provides a data preprocessing system and a method thereof for a supply chain planning system, wherein the data preprocessing system acquires a set of characteristic value data to be processed; carrying out association feature analysis based on a subset on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data; taking a certain inter-eigenvalue data associated eigenvector in the sequence of the inter-eigenvalue data associated eigenvector as a query eigenvector, and calculating a data distribution difference semantic measurement coefficient between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of the inter-eigenvalue data associated eigenvector; and determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the data distribution difference semantic measurement coefficient. In this way, the supply chain planning system may be aided in better understanding and managing the data, thereby improving the accuracy and efficiency of the decisions.
The embodiment of the invention also provides a data preprocessing method for the supply chain planning system, which comprises the following steps: acquiring a set of feature value data to be processed; carrying out association feature analysis based on a subset on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data; taking a certain inter-eigenvalue data associated eigenvector in the sequence of the inter-eigenvalue data associated eigenvector as a query eigenvector, and calculating a data distribution difference semantic measurement coefficient between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of the inter-eigenvalue data associated eigenvector; and determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the data distribution difference semantic measurement coefficient.
In the data preprocessing method for the supply chain planning system, the characteristic value data to be processed is the latest restocking period.
In the above data preprocessing method for a supply chain planning system, performing a subset-based correlation feature analysis on the set of feature value data to be processed to obtain a sequence of correlation feature vectors among the feature value data, including: performing data preprocessing on the set of the characteristic value data to be processed to obtain a sequence of characteristic value data input vectors; and extracting features of the sequence of the feature value data input vectors by using a deep learning network model to obtain a sequence of the correlation feature vectors among the feature value data.
In the above data preprocessing method for a supply chain planning system, the data preprocessing is performed on the set of feature value data to be processed to obtain a sequence of feature value data input vectors, including: carrying out subset division on the set of the characteristic value data to be processed to obtain a set of the subset of the characteristic value data to be processed; and arranging the subsets of the characteristic value data to be processed in the set of the subsets of the characteristic value data to be processed according to the characteristic value data sample dimension to obtain a sequence of the characteristic value data input vector.
In the data preprocessing method for the supply chain planning system, the deep learning network model is a feature value data subset mode feature extractor based on a one-dimensional convolution layer.
In the above data preprocessing method for a supply chain planning system, performing feature extraction on the sequence of feature value data input vectors by using a deep learning network model to obtain the sequence of associated feature vectors among the feature value data, including: and passing each eigenvalue data input vector in the sequence of eigenvalue data input vectors through the eigenvalue data subset mode feature extractor based on the one-dimensional convolution layer to obtain the sequence of correlation eigenvectors among the eigenvalue data.
In the above data preprocessing method for a supply chain planning system, using a certain inter-eigenvalue data associated eigenvector in the sequence of inter-eigenvalue data associated eigenvectors as a query eigenvector, calculating a data distribution difference semantic measurement coefficient between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of inter-eigenvalue data associated eigenvectors, including: calculating the data distribution difference semantic measurement coefficient between the associated feature vector among certain feature value data and the associated feature vector among all other feature value data in the sequence of the associated feature vector among the feature value data according to the following data distribution difference semantic measurement formula; the data distribution difference semantic measurement formula is as follows: ; wherein, Is the first in the sequence of the correlation feature vector between the feature value dataA correlation feature vector among the feature value data, namely the correlation feature vector among the feature value data,Is the first in the sequence of the correlation feature vector between the feature value dataThe feature vectors are associated between the feature value data,Representing the 1-norm of the feature vector,For the length-1 of the sequence of associated feature vectors between the feature value data,For the representation of the sequence of associated feature vectors between the feature value data,Representing the data distribution difference semantic metric coefficient.
In the above data preprocessing method for a supply chain planning system, determining whether abnormal data exists in the subset of the to-be-processed eigenvalue data corresponding to the associated eigenvalue vector among the certain eigenvalue data based on the data distribution difference semantic metric coefficient includes: and determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the comparison between the data distribution difference semantic measurement coefficient and a preset threshold value.
The data preprocessing method for the supply chain planning system further comprises the training step of: training the feature value data subset mode feature extractor based on the one-dimensional convolution layer; wherein the training step comprises: acquiring a set of training to-be-processed characteristic value data; carrying out subset division on the set of the training to-be-processed characteristic value data to obtain a set of the subset of the training to-be-processed characteristic value data; arranging the subsets of the training to-be-processed characteristic value data in the subset set of the training to-be-processed characteristic value data according to the characteristic value data sample dimension to obtain a sequence of training characteristic value data input vectors; each training characteristic value data input vector in the sequence of training characteristic value data input vectors passes through the characteristic value data subset mode characteristic extractor based on the one-dimensional convolution layer to obtain a sequence of correlation characteristic vectors among training characteristic value data; taking a certain training characteristic value data associated characteristic vector in the sequence of training characteristic value data associated characteristic vectors as a query characteristic vector, and calculating training data distribution difference semantic measurement coefficients between the query characteristic vector and all other training characteristic value data associated characteristic vectors in the sequence of training characteristic value data associated characteristic vectors; acquiring the true value of the data distribution difference semantic measurement coefficient between the associated feature vector among certain training feature value data and all other associated feature vectors among the training feature value data in the sequence of the associated feature vectors among the training feature value data; calculating a cross entropy function value between the training data distribution difference semantic measurement coefficient and a true value of the data distribution difference semantic measurement coefficient; training the feature extractor based on the feature value data subset mode of the one-dimensional convolution layer by taking the cross entropy function value as a difference loss function value, wherein in each iteration of the training, the sequence of the associated feature vectors among the training feature value data is optimized.
The embodiment of the invention also provides a data preprocessing system for a supply chain planning system, which comprises the following steps: the data acquisition module is used for acquiring a set of characteristic value data to be processed; the associated feature analysis module is used for carrying out associated feature analysis based on subsets on the set of the feature value data to be processed so as to obtain a sequence of associated feature vectors among the feature value data; the semantic measurement coefficient calculation module is used for calculating the data distribution difference semantic measurement coefficient between the semantic measurement coefficient and all other associated feature vectors in the sequence of the associated feature vectors between the feature values by taking the associated feature vector between certain feature values in the sequence of the associated feature vectors between the feature values as the query feature vector; and the abnormal data judging module is used for determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the data distribution difference semantic measurement coefficient.
Compared with the prior art, the data preprocessing system and the method for the supply chain planning system acquire the set of the characteristic value data to be processed; carrying out association feature analysis based on a subset on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data; taking a certain inter-eigenvalue data associated eigenvector in the sequence of the inter-eigenvalue data associated eigenvector as a query eigenvector, and calculating a data distribution difference semantic measurement coefficient between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of the inter-eigenvalue data associated eigenvector; and determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the data distribution difference semantic measurement coefficient. In this way, the supply chain planning system may be aided in better understanding and managing the data, thereby improving the accuracy and efficiency of the decisions.
Drawings
In order to more clearly illustrate the embodiments of the 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, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
In the drawings: FIG. 1 is a flow chart of a method for data preprocessing for a supply chain planning system according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a system architecture of a data preprocessing method for a supply chain planning system according to an embodiment of the present invention.
FIG. 3 is a block diagram of a data preprocessing system for a supply chain planning system provided in an embodiment of the present invention.
Fig. 4 is an application scenario diagram of a data preprocessing method for a supply chain planning system according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings. The exemplary embodiments of the present invention and their descriptions herein are for the purpose of explaining the present invention, but are not to be construed as limiting the invention.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
A flowchart is used in the present invention to describe the operations performed by a system according to embodiments of the present invention. It should be understood that the preceding or following operations are not necessarily performed in order precisely. Rather, the various steps may be processed in reverse order or simultaneously, as desired. Also, other operations may be added to or removed from these processes.
Various exemplary embodiments, features and aspects of the disclosure will be described in detail below with reference to the drawings. In the drawings, like reference numbers indicate identical or functionally similar elements. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
The word "exemplary" is used herein to mean "serving as an example, embodiment, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.
A supply chain planning system is a software system dedicated to managing various resources and activities in a supply chain network, the main objectives of which are to help businesses increase efficiency, reduce costs, increase profits, and meet customer needs, and one of the core functions of such a system is to generate optimal production, procurement, inventory, and transportation plans based on forecasted and actual needs. To achieve this, the supply chain planning system needs to process a large amount of data, including main data and business data.
Master data refers to data describing entities and relationships in a supply chain network, such as materials, factories, warehouses, suppliers, customers, etc., that provide information about the structure and basic features of the supply chain network, which is the basis of the supply chain planning system. The business data describes the operation conditions in the supply chain network, including information such as orders, inventory, production, purchase, transportation, etc., and the data is the basis for the supply chain planning system to generate plans.
The quality of the data is critical to the performance and effectiveness of the supply chain planning system, and the high quality data can improve the accuracy and reliability of the system, thereby ensuring that the generated plan more conforms to the actual situation. Thus, ensuring the accuracy, integrity and timeliness of the data is critical to the proper operation of the supply chain planning system.
In addition to processing data, supply chain planning systems often have other functions, such as demand forecasting, inventory optimization, haul route optimization, etc., to comprehensively manage and optimize the entire supply chain network, the combination of which can help businesses achieve more efficient operation and more optimal resource utilization, thereby increasing competitiveness and meeting customer demands.
In practice, supply chain planning systems do face a number of challenges in terms of data quality. Data non-norms, inconsistencies, imperfections, and inaccuracies may negatively impact the performance and effectiveness of the system, thereby affecting the operational efficiency and customer satisfaction of the enterprise, and in order to address these issues, a data preprocessing system and method for a supply chain planning system may be desired.
Such a data preprocessing system may include a series of steps of data cleansing, data integration, data conversion, and data quality assessment. First, data cleansing may be used to identify and correct errors, deletions, or inconsistencies in data to ensure accuracy and integrity of the data. Second, data integration may integrate data from different sources into a unified data store to eliminate duplicate data and ensure data consistency. Data conversion may be used to convert data into a format and structure suitable for use by a supply chain planning system. Finally, the data quality assessment may assess the data quality through various metrics and methods to discover and resolve potential problems in time.
In addition to data preprocessing systems, continuous monitoring and improvement of data quality using data quality management tools and techniques that can help enterprises identify the sources of data quality problems and take corresponding measures to improve, thereby ensuring that supply chain planning systems always use high quality data for planning and decision making, can also be considered.
Therefore, the data preprocessing system and the method thereof which are specially aimed at the supply chain planning system are established, and the data quality management tool and the technology are adopted, so that the data quality challenges faced by the supply chain planning system can be effectively met, the accuracy and the reliability of the system are improved, and better support is provided for the operation efficiency and the customer satisfaction of enterprises.
In the present application, a data preprocessing system for a supply chain planning system and a method thereof are provided, which can implement rapid batch analysis of problem data: first, rapid batch analysis of problem data can help development teams identify data quality problems in a short period of time. Through automated data analysis tools, development teams can quickly identify problems such as non-norms, inconsistencies, imperfections, and the like in the data, so as to correct in time. The automatic analysis mode can greatly improve the development efficiency and reduce the time and cost required by manual analysis. This advantage is even more pronounced especially for companies where the supply chain network is more complex. Secondly, the rapid batch analysis of problem data can improve the stability and reliability of the system. By analyzing the problem data in the system main data and the service data, the potential problems possibly causing the false planning can be rapidly identified, and the repair can be timely carried out, so that the stability and the reliability of the system are ensured. The automatic data analysis mode can help a business team and an operation and maintenance team to better know the operation condition of the system, and timely identify and solve possible problems. In the setting of cloud software, related programs can be executed in the background at regular intervals, and related personnel are informed of problem data in the form of logs or emails so as to obtain repair as soon as possible. Finally, the quick batch analysis of the problem data can improve customer satisfaction. By analyzing the problem data in the system main data and the service data, the data quality problem of the client can be found and solved in time, so that the satisfaction degree of the client is improved.
In one embodiment of the present invention, FIG. 1 is a flow chart of a method for data preprocessing for a supply chain planning system provided in an embodiment of the present invention. Fig. 2 is a schematic diagram of a system architecture of a data preprocessing method for a supply chain planning system according to an embodiment of the present invention. As shown in fig. 1 and 2, a data preprocessing method for a supply chain planning system according to an embodiment of the present invention includes: 110, acquiring a set of feature value data to be processed; 120, carrying out association feature analysis based on a subset on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data; 130, calculating a data distribution difference semantic measurement coefficient between the query feature vector and all other feature value data associated feature vectors in the sequence of feature value data associated feature vectors by taking a certain feature value data associated feature vector in the sequence of feature value data associated feature vectors as the query feature vector; 140, determining whether abnormal data exists in the subset of the to-be-processed eigenvalue data corresponding to the correlation eigenvector among the certain eigenvalue data based on the data distribution difference semantic measurement coefficient.
In said step 110 a set of feature value data to be processed is acquired, at which step it is ensured that a complete set of feature value data to be processed is acquired and that the quality and integrity of the data is reliable. The complete to-be-processed characteristic value data set is obtained to be the basis of subsequent analysis, so that the comprehensiveness and accuracy of analysis are ensured.
In the step 120, a subset-based correlation feature analysis is performed on the set of feature value data to be processed to obtain a sequence of correlation feature vectors between feature value data, and when the correlation feature analysis is performed, a suitable method is selected to identify the correlation between feature value data, for example, a correlation analysis, a cluster analysis, or the like may be used. Through correlation feature analysis, a correlation mode between feature value data can be found, and the mutual influence between the internal structure and the features of the data is facilitated to be understood.
In the step 130, a certain inter-eigenvalue data associated eigenvector in the sequence of inter-eigenvalue data associated eigenvectors is used as a query eigenvector, a semantic measurement coefficient of the data distribution difference between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of inter-eigenvalue data associated eigenvector is calculated, and it is critical to select a proper measurement method, so as to ensure that the selected method can accurately measure the data distribution difference between the eigenvectors. The data distribution difference between the feature vectors can be quantified by calculating the semantic measurement coefficient of the data distribution difference, so that a basis is provided for the subsequent abnormal data identification.
In the step 140, based on the data distribution difference semantic metric coefficient, it is determined whether there is abnormal data in the subset of the to-be-processed feature value data corresponding to the associated feature vector between the certain feature value data, and it is critical to select a suitable abnormal data identification method, for example, a statistical method, a machine learning method, or a data distribution-based method may be used to identify abnormal data. By identifying outlier data within a subset of the eigenvalue data, it may be helpful to discover outlier patterns or outliers in the data, providing basis for further processing.
Aiming at the technical problems, the technical conception of the application is as follows: and carrying out subset division, feature extraction and anomaly detection on the feature value data in the supply chain planning system by utilizing an intelligent algorithm so as to identify and reject the anomaly data, thereby improving the data quality. Taking the latest restocking period as an example, it is desirable to sub-divide the set and determine whether abnormal data exists in a specific subset based on the data distribution joint difference between the subsets. In addition, after determining that the abnormal data exists in the specific subset, the specific subset is regarded as a set to be subjected to subset division, and the positions of the abnormal data are determined in a cyclic reciprocation mode.
Based on this, in the technical scheme of the application, first, a set of feature value data to be processed is acquired. The set of feature value data to be processed refers to a set of feature value data which needs to be subjected to data preprocessing in a supply chain planning system. In a supply chain planning system, the eigenvalue data typically includes various attributes and indices related to materials, factories, warehouses, suppliers, customers, and the like. Such data is used to predict and plan supply chain activities such as production planning, procurement planning, inventory management, transportation planning, and the like. In particular, the set of eigenvalue data to be processed may be a data table or a data set, wherein each row represents one data sample and each column represents one eigenvalue data. For example, for material related characteristic value data, properties such as material code, material description, inventory, demand, replenishment period, etc. may be included. The plant-related feature value data may include attributes such as plant code, plant location, throughput, and capacity utilization.
In an embodiment of the present application, the feature value data to be processed is a latest restocking period. Here, the latest restocking period (LATEST REPLENISHMENT DATE) refers to the deadline by which restocking or restocking of a particular material or product needs to be completed before that date in supply chain management. The latest restocking period is determined based on factors such as supply chain planning and demand forecast to ensure that the supply chain can meet customer demand on time and avoid stock out. More specifically, the latest restocking period is typically determined based on factors of demand planning, delivery time, transit time, buffer inventory, etc. of the supply chain. It takes into account the time requirements of each link in the supply chain, including the delivery time, the transit time, and the internal processing time of the supplier. By setting the latest restocking period, the supply chain manager can schedule restocking activities according to demand forecast and delivery time to ensure that inventory can meet customer demand in time while avoiding excessive inventory backlog.
In one embodiment of the present application, performing a subset-based correlation feature analysis on the set of feature value data to be processed to obtain a sequence of correlation feature vectors among the feature value data, including: performing data preprocessing on the set of the characteristic value data to be processed to obtain a sequence of characteristic value data input vectors; and extracting features of the sequence of the feature value data input vectors by using a deep learning network model to obtain a sequence of the correlation feature vectors among the feature value data.
Then, carrying out subset division on the set of the characteristic value data to be processed to obtain a set of the subset of the characteristic value data to be processed; and arranging the subsets of the feature value data to be processed in the set of the subsets of the feature value data to be processed according to the feature value data sample dimension to obtain a sequence of feature value data input vectors. Here, considering that the set of the feature value data to be processed is very huge, dividing it into a plurality of subsets may improve the efficiency of data processing. By processing each subset in parallel, processing time can be reduced and computational performance can be improved. And then, arranging the subsets of the to-be-processed characteristic value data in the set of the subsets of the to-be-processed characteristic value data according to the characteristic value data sample dimension, so that the structural information of the data can be better utilized. That is, for each eigenvalue data sample, it is arranged to form individual eigenvalue data input vectors that can be more conveniently read and analyzed by subsequent models.
In a specific embodiment of the present application, the data preprocessing is performed on the set of feature value data to be processed to obtain a sequence of feature value data input vectors, including: carrying out subset division on the set of the characteristic value data to be processed to obtain a set of the subset of the characteristic value data to be processed; and arranging the subsets of the characteristic value data to be processed in the set of the subsets of the characteristic value data to be processed according to the characteristic value data sample dimension to obtain a sequence of the characteristic value data input vector.
And then, each characteristic value data input vector in the sequence of characteristic value data input vectors passes through a characteristic extractor based on a characteristic value data subset mode of a one-dimensional convolution layer to obtain a sequence of associated characteristic vectors among characteristic value data. That is, the feature value data subset pattern feature extractor is constructed with a one-dimensional convolution layer to capture local patterns and associated information in the feature value data.
The deep learning network model is a feature extractor of a feature value data subset mode based on a one-dimensional convolution layer.
In a specific embodiment of the present application, feature extraction is performed on the sequence of feature value data input vectors by using a deep learning network model to obtain a sequence of associated feature vectors among the feature value data, including: and passing each eigenvalue data input vector in the sequence of eigenvalue data input vectors through the eigenvalue data subset mode feature extractor based on the one-dimensional convolution layer to obtain the sequence of correlation eigenvectors among the eigenvalue data.
More specifically, the one-dimensional convolution layer may extract the local pattern in each eigenvalue data input vector by means of a sliding window. The convolution check performs convolution operation on each window, so that a local characteristic mode, such as an ascending trend, a descending trend or a periodic mode, is captured. By extracting the local patterns, the characteristics and rules of the data can be better understood. In addition, there may be a certain correlation in the feature value data. The relative position and sequence relation between the characteristic value data can be captured through the one-dimensional convolution layer. The convolution kernel, when slid over the eigenvalue data, can identify patterns of relationship between the eigenvalue data, such as where certain data are concurrent or interdependent. This helps extract interaction and correlation information between features. In this way, the feature extractor of the feature value data subset mode based on the one-dimensional convolution layer can extract local modes and associated information in the feature value data, thereby helping to reveal abnormal distribution in the set of the feature value data to be processed.
And then, taking a certain inter-eigenvalue data associated eigenvector in the sequence of the inter-eigenvalue data associated eigenvector as a query eigenvector, and calculating the data distribution difference semantic measurement coefficient between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of the inter-eigenvalue data associated eigenvector. The data distribution difference semantic measurement coefficient is used for measuring the data characteristic distribution in the corresponding subset expressed by the related characteristic vector among certain characteristic value data, and the degree of difference among the data characteristic distribution of the whole subset expressed by the related characteristic vector among other all characteristic value data is used as an important basis for judging whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the related characteristic vector among certain characteristic value data.
In a specific embodiment of the present application, a query feature vector is a feature value inter-data correlation feature vector in the sequence of feature value inter-data correlation feature vectors, and a data distribution difference semantic measurement coefficient between the query feature vector and all other feature value inter-data correlation feature vectors in the sequence of feature value inter-data correlation feature vectors is calculated, including: calculating the data distribution difference semantic measurement coefficient between the associated feature vector among certain feature value data and the associated feature vector among all other feature value data in the sequence of the associated feature vector among the feature value data according to the following data distribution difference semantic measurement formula; the data distribution difference semantic measurement formula is as follows: ; wherein, Is the first in the sequence of the correlation feature vector between the feature value dataA correlation feature vector among the feature value data, namely the correlation feature vector among the feature value data,Is the first in the sequence of the correlation feature vector between the feature value dataThe feature vectors are associated between the feature value data,Representing the 1-norm of the feature vector,For the length-1 of the sequence of associated feature vectors between the feature value data,For the representation of the sequence of associated feature vectors between the feature value data,Representing the data distribution difference semantic metric coefficient.
Further, based on the comparison between the data distribution difference semantic measurement coefficient and a preset threshold value, whether abnormal data exist in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data is determined. It should be noted that the threshold is set according to the specific application and requirement. It may be determined based on domain knowledge, historical experience, or statistical analysis, among other methods. The selection of the threshold value needs to comprehensively consider the accuracy and the error rate of anomaly detection.
In a specific embodiment of the present application, based on the data distribution difference semantic metric coefficient, determining whether abnormal data exists in the subset of the to-be-processed feature value data corresponding to the associated feature vector between the certain feature value data includes: and determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the comparison between the data distribution difference semantic measurement coefficient and a preset threshold value.
In one embodiment of the present application, the data preprocessing method for the supply chain planning system further comprises a training step of: training the feature value data subset mode feature extractor based on the one-dimensional convolution layer; wherein the training step comprises: acquiring a set of training to-be-processed characteristic value data; carrying out subset division on the set of the training to-be-processed characteristic value data to obtain a set of the subset of the training to-be-processed characteristic value data; arranging the subsets of the training to-be-processed characteristic value data in the subset set of the training to-be-processed characteristic value data according to the characteristic value data sample dimension to obtain a sequence of training characteristic value data input vectors; each training characteristic value data input vector in the sequence of training characteristic value data input vectors passes through the characteristic value data subset mode characteristic extractor based on the one-dimensional convolution layer to obtain a sequence of correlation characteristic vectors among training characteristic value data; taking a certain training characteristic value data associated characteristic vector in the sequence of training characteristic value data associated characteristic vectors as a query characteristic vector, and calculating training data distribution difference semantic measurement coefficients between the query characteristic vector and all other training characteristic value data associated characteristic vectors in the sequence of training characteristic value data associated characteristic vectors; acquiring the true value of the data distribution difference semantic measurement coefficient between the associated feature vector among certain training feature value data and all other associated feature vectors among the training feature value data in the sequence of the associated feature vectors among the training feature value data; calculating a cross entropy function value between the training data distribution difference semantic measurement coefficient and a true value of the data distribution difference semantic measurement coefficient; training the feature extractor based on the feature value data subset mode of the one-dimensional convolution layer by taking the cross entropy function value as a difference loss function value, wherein in each iteration of the training, the sequence of the associated feature vectors among the training feature value data is optimized.
In the technical scheme of the application, each training eigenvalue data inter-associated eigenvector in the sequence of training eigenvalue data inter-associated eigenvector expresses the local associated eigenvalue of data in a subset of the training eigenvalue data to be processed, therefore, the associated eigenvector among a certain training eigenvalue data in the sequence of training eigenvalue inter-associated eigenvalue data is used as a query eigenvector, when the data distribution difference semantic measurement coefficient between the training eigenvalue data inter-associated eigenvector and all other training eigenvalue data inter-associated eigenvector in the sequence of training eigenvalue inter-associated eigenvector is calculated, a vector set formed by the other all training eigenvalue data inter-associated eigenvector can cause the associated eigenvalue distribution information significance of the associated eigenvector among the training eigenvalue data due to the associated eigenvalue distribution difference of the source data subset of the associated eigenvalue data, so that the associated eigenvalue data associated eigenvector is used as a set and is difficult to be focused on the significant local distribution of the feature in the process, thereby influencing the calculation accuracy of the data distribution difference semantic measurement coefficient.
Based on this, in each iteration of the training of the present application, the sequence of associated feature vectors between the training feature value data is optimized, for example, when a difference loss function between the inferred data distribution difference semantic metric coefficient and the actual data distribution difference semantic metric coefficient is back-propagated through the sequence of associated feature vectors between the training feature value data.
In a specific embodiment of the present application, in each iteration of the training, optimizing the sequence of the associated feature vectors between the training feature value data includes: firstly, cascading the sequence of the correlation feature vectors among the training feature value data into the correlation cascade feature vectors among the training feature value data, calculating the self-correlation matrix of the correlation cascade feature vectors among the training feature value data and the transposition of the self-correlation matrix, and calculating the first self-correlation matrixAnd (d)Inner product of row vectors as the first of the weight matrixMatrix values of the positions are used for obtaining a weight matrix, then, after the weight matrix is multiplied by the self-correlation matrix, matrix-vector multiplication is further carried out on the weight matrix and the training eigenvalue data correlation cascade eigenvector to obtain a correction vector, finally, the correction vector and the training eigenvalue data correlation cascade eigenvector point multiplication are carried out to obtain an optimized training eigenvalue data correlation cascade eigenvector, and the optimized training eigenvalue data correlation cascade eigenvector is restored to be a sequence of optimized training eigenvalue data correlation eigenvector.
In this way, based on the self-association dimension of the associated cascade feature vector among the training feature value data, which is the object to be modulated, the association expansion is performed based on the space sub-dimension complexity of the high-dimensional feature space of the feature distribution, so that the resolvable association dimension offset of the associated cascade feature vector among the training feature value data is introduced into the heterogeneous association embedding space, the association self-consistent relation is enhanced through the joint fine tuning of the resolvable dimension set represented by the heterogeneous association embedding space, so that the predetermined feature focusing distribution of the sequence of the associated feature vector among the training feature value data in the associated target semantic measurement space is improved, and the calculation accuracy of the data distribution difference semantic measurement coefficient between the associated feature vector among the training feature value data and the associated feature vector among all other training feature value data in the sequence of the associated feature vector among the training feature value data is improved.
In summary, the data preprocessing method for the supply chain planning system according to the embodiment of the invention is explained, which utilizes an intelligent algorithm to perform subset division, feature extraction and anomaly detection on the feature value data in the supply chain planning system so as to identify and reject the anomaly data, thereby improving the data quality. Taking the latest restocking period as an example, it is desirable to sub-divide the set and determine whether abnormal data exists in a specific subset based on the data distribution joint difference between the subsets. In addition, after determining that the abnormal data exists in the specific subset, the specific subset is regarded as a set to be subjected to subset division, and the positions of the abnormal data are determined in a cyclic reciprocation mode.
In one embodiment of the present invention, FIG. 3 is a block diagram of a data preprocessing system for a supply chain planning system provided in an embodiment of the present invention. As shown in fig. 3, a data preprocessing system 200 for a supply chain planning system according to an embodiment of the present invention includes: a data acquisition module 210, configured to acquire a set of feature value data to be processed; the correlation feature analysis module 220 is configured to perform a subset-based correlation feature analysis on the set of feature value data to be processed to obtain a sequence of correlation feature vectors between the feature value data; the semantic measurement coefficient calculating module 230 is configured to calculate a data distribution difference semantic measurement coefficient between the semantic measurement coefficient and all other inter-feature value data associated feature vectors in the sequence of inter-feature value data associated feature vectors, with the inter-feature value data associated feature vector in the sequence of inter-feature value data associated feature vectors as a query feature vector; the abnormal data judging module 240 is configured to determine, based on the data distribution difference semantic metric coefficient, whether abnormal data exists in the subset of the to-be-processed feature value data corresponding to the associated feature vector between the certain feature value data.
Here, it will be understood by those skilled in the art that the specific functions and operations of the respective units and modules in the above-described data preprocessing system for a supply chain planning system have been described in detail in the above description of the data preprocessing method for a supply chain planning system with reference to fig. 1 to 2, and thus, repetitive descriptions thereof will be omitted.
As described above, the data preprocessing system 200 for a supply chain planning system according to an embodiment of the present invention may be implemented in various terminal devices, such as a server or the like for data preprocessing of a supply chain planning system. In one example, the data preprocessing system 200 for the supply chain planning system according to an embodiment of the present invention may be integrated into the terminal device as one software module and/or hardware module. For example, the data preprocessing system 200 for the supply chain planning system may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the data preprocessing system 200 for the supply chain planning system can equally be one of the numerous hardware modules of the terminal device.
Alternatively, in another example, the data preprocessing system 200 for the supply chain planning system and the terminal device may be separate devices, and the data preprocessing system 200 for the supply chain planning system may be connected to the terminal device through a wired and/or wireless network and transmit interactive information in a contracted data format.
Fig. 4 is an application scenario diagram of a data preprocessing method for a supply chain planning system according to an embodiment of the present invention. As shown in fig. 4, in this application scenario, first, a set of feature value data to be processed is acquired (e.g., C as illustrated in fig. 4); then, the acquired set of feature value data to be processed is input into a server (e.g., S as illustrated in fig. 4) deployed with a data preprocessing algorithm for a supply chain planning system, wherein the server is capable of processing the set of feature value data to be processed based on the data preprocessing algorithm for the supply chain planning system to determine whether or not there is abnormal data within a subset of the feature value data to be processed corresponding to an associated feature vector among the certain feature value data.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (7)
1. A method of data preprocessing for a supply chain planning system, comprising:
acquiring a set of feature value data to be processed;
carrying out association feature analysis based on a subset on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data;
taking a certain inter-eigenvalue data associated eigenvector in the sequence of the inter-eigenvalue data associated eigenvector as a query eigenvector, and calculating a data distribution difference semantic measurement coefficient between the query eigenvector and all other inter-eigenvalue data associated eigenvectors in the sequence of the inter-eigenvalue data associated eigenvector;
determining whether abnormal data exists in a subset of the to-be-processed characteristic value data corresponding to the associated characteristic vector among the certain characteristic value data based on the data distribution difference semantic measurement coefficient;
The method for processing the feature value data comprises the steps of carrying out association feature analysis based on subsets on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data, and comprises the following steps:
Performing data preprocessing on the set of the characteristic value data to be processed to obtain a sequence of characteristic value data input vectors;
Extracting features of the sequence of the feature value data input vectors by using a deep learning network model to obtain a sequence of the correlation feature vectors among the feature value data;
the data preprocessing is performed on the set of feature value data to be processed to obtain a sequence of feature value data input vectors, and the data preprocessing comprises the following steps:
carrying out subset division on the set of the characteristic value data to be processed to obtain a set of the subset of the characteristic value data to be processed;
arranging the subsets of the feature value data to be processed in the set of the subsets of the feature value data to be processed according to the feature value data sample dimension to obtain a sequence of the feature value data input vector;
the method for calculating the semantic measurement coefficient of the data distribution difference between the correlation feature vector and all other feature value data in the sequence of the correlation feature vector between the feature value data by taking the correlation feature vector between certain feature value data in the sequence of the correlation feature vector between the feature value data as a query feature vector comprises the following steps:
Calculating the data distribution difference semantic measurement coefficient between the associated feature vector among certain feature value data and the associated feature vector among all other feature value data in the sequence of the associated feature vector among the feature value data according to the following data distribution difference semantic measurement formula; the data distribution difference semantic measurement formula is as follows:
；
wherein, Is the first in the sequence of the correlation feature vector between the feature value dataA correlation feature vector among the feature value data, namely the correlation feature vector among the feature value data,Is the first in the sequence of the correlation feature vector between the feature value dataThe feature vectors are associated between the feature value data,Representing the 1-norm of the feature vector,For the length-1 of the sequence of associated feature vectors between the feature value data,For the representation of the sequence of associated feature vectors between the feature value data,Representing the data distribution difference semantic metric coefficient.
2. The method of claim 1, wherein the characteristic value data to be processed is a latest restocking period.
3. The method of data preprocessing for a supply chain planning system of claim 2, wherein said deep learning network model is a feature value data subset pattern feature extractor based on a one-dimensional convolution layer.
4. The method of claim 3, wherein performing feature extraction on the sequence of eigenvalue data input vectors using a deep learning network model to obtain a sequence of associated eigenvectors among the eigenvalue data comprises:
And passing each eigenvalue data input vector in the sequence of eigenvalue data input vectors through the eigenvalue data subset mode feature extractor based on the one-dimensional convolution layer to obtain the sequence of correlation eigenvectors among the eigenvalue data.
5. The method for data preprocessing for a supply chain planning system according to claim 4, wherein determining whether abnormal data exists in the subset of the to-be-processed eigenvalue data corresponding to the certain inter-eigenvalue data associated eigenvector based on the data distribution difference semantic metric coefficient comprises:
And determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the comparison between the data distribution difference semantic measurement coefficient and a preset threshold value.
6. The method for data preprocessing for a supply chain planning system according to claim 5, further comprising a training step of: training the feature value data subset mode feature extractor based on the one-dimensional convolution layer;
Wherein the training step comprises:
acquiring a set of training to-be-processed characteristic value data;
Carrying out subset division on the set of the training to-be-processed characteristic value data to obtain a set of the subset of the training to-be-processed characteristic value data;
Arranging the subsets of the training to-be-processed characteristic value data in the subset set of the training to-be-processed characteristic value data according to the characteristic value data sample dimension to obtain a sequence of training characteristic value data input vectors;
each training characteristic value data input vector in the sequence of training characteristic value data input vectors passes through the characteristic value data subset mode characteristic extractor based on the one-dimensional convolution layer to obtain a sequence of correlation characteristic vectors among training characteristic value data;
taking a certain training characteristic value data associated characteristic vector in the sequence of training characteristic value data associated characteristic vectors as a query characteristic vector, and calculating training data distribution difference semantic measurement coefficients between the query characteristic vector and all other training characteristic value data associated characteristic vectors in the sequence of training characteristic value data associated characteristic vectors;
Acquiring the true value of the data distribution difference semantic measurement coefficient between the associated feature vector among certain training feature value data and all other associated feature vectors among the training feature value data in the sequence of the associated feature vectors among the training feature value data;
calculating a cross entropy function value between the training data distribution difference semantic measurement coefficient and a true value of the data distribution difference semantic measurement coefficient;
Training the feature extractor based on the feature value data subset mode of the one-dimensional convolution layer by taking the cross entropy function value as a difference loss function value, wherein in each iteration of the training, the sequence of the associated feature vectors among the training feature value data is optimized.
7. A data preprocessing system for a supply chain planning system, comprising:
the data acquisition module is used for acquiring a set of characteristic value data to be processed;
The associated feature analysis module is used for carrying out associated feature analysis based on subsets on the set of the feature value data to be processed so as to obtain a sequence of associated feature vectors among the feature value data;
The semantic measurement coefficient calculation module is used for calculating the data distribution difference semantic measurement coefficient between the semantic measurement coefficient and all other associated feature vectors in the sequence of the associated feature vectors between the feature values by taking the associated feature vector between certain feature values in the sequence of the associated feature vectors between the feature values as the query feature vector;
The abnormal data judging module is used for determining whether abnormal data exists in the subset of the to-be-processed characteristic value data corresponding to the correlation characteristic vector among the certain characteristic value data based on the data distribution difference semantic measurement coefficient;
The method for processing the feature value data comprises the steps of carrying out association feature analysis based on subsets on the set of the feature value data to be processed to obtain a sequence of association feature vectors among the feature value data, and comprises the following steps:
Performing data preprocessing on the set of the characteristic value data to be processed to obtain a sequence of characteristic value data input vectors;
Extracting features of the sequence of the feature value data input vectors by using a deep learning network model to obtain a sequence of the correlation feature vectors among the feature value data;
the data preprocessing is performed on the set of feature value data to be processed to obtain a sequence of feature value data input vectors, and the data preprocessing comprises the following steps:
carrying out subset division on the set of the characteristic value data to be processed to obtain a set of the subset of the characteristic value data to be processed;
arranging the subsets of the feature value data to be processed in the set of the subsets of the feature value data to be processed according to the feature value data sample dimension to obtain a sequence of the feature value data input vector;
the method for calculating the semantic measurement coefficient of the data distribution difference between the correlation feature vector and all other feature value data in the sequence of the correlation feature vector between the feature value data by taking the correlation feature vector between certain feature value data in the sequence of the correlation feature vector between the feature value data as a query feature vector comprises the following steps:
Calculating the data distribution difference semantic measurement coefficient between the associated feature vector among certain feature value data and the associated feature vector among all other feature value data in the sequence of the associated feature vector among the feature value data according to the following data distribution difference semantic measurement formula; the data distribution difference semantic measurement formula is as follows:
；
wherein, Is the first in the sequence of the correlation feature vector between the feature value dataA correlation feature vector among the feature value data, namely the correlation feature vector among the feature value data,Is the first in the sequence of the correlation feature vector between the feature value dataThe feature vectors are associated between the feature value data,Representing the 1-norm of the feature vector,For the length-1 of the sequence of associated feature vectors between the feature value data,For the representation of the sequence of associated feature vectors between the feature value data,Representing the data distribution difference semantic metric coefficient.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410150627 | 2024-02-02 | ||
| CN2024101506279 | 2024-02-02 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118395135A CN118395135A (en) | 2024-07-26 |
| CN118395135B true CN118395135B (en) | 2024-09-03 |
Family
ID=92006400
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410866238.6A Active CN118395135B (en) | 2024-02-02 | 2024-07-01 | Data preprocessing system and method for supply chain planning system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118395135B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109670549A (en) * | 2018-12-20 | 2019-04-23 | 华润电力技术研究院有限公司 | The data screening method, apparatus and computer equipment of fired power generating unit |
| CN116956190A (en) * | 2023-07-17 | 2023-10-27 | 广州大学 | Malicious information detection method and device, electronic equipment and storage medium |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4513796B2 (en) * | 2006-10-12 | 2010-07-28 | パナソニック電工株式会社 | Abnormality monitoring device |
| JP6076751B2 (en) * | 2013-01-22 | 2017-02-08 | 株式会社日立製作所 | Abnormality diagnosis method and apparatus |
| US10635519B1 (en) * | 2017-11-30 | 2020-04-28 | Uptake Technologies, Inc. | Systems and methods for detecting and remedying software anomalies |
| CN114494242A (en) * | 2022-02-21 | 2022-05-13 | 平安科技（深圳）有限公司 | Time series data detection method, device, equipment and computer storage medium |
| CN115840774A (en) * | 2022-11-25 | 2023-03-24 | 北京航空航天大学杭州创新研究院 | Multi-element time sequence abnormity detection method and device, computer equipment and storage medium |
| CN116340039A (en) * | 2023-03-03 | 2023-06-27 | 同济大学 | Log anomaly detection method based on pretrained BERT sentence vector and Informar-encoder |
| CN117476214A (en) * | 2023-11-10 | 2024-01-30 | 郑州蓝博电子技术有限公司 | Data management method and system based on hospital information |
-
2024
- 2024-07-01 CN CN202410866238.6A patent/CN118395135B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109670549A (en) * | 2018-12-20 | 2019-04-23 | 华润电力技术研究院有限公司 | The data screening method, apparatus and computer equipment of fired power generating unit |
| CN116956190A (en) * | 2023-07-17 | 2023-10-27 | 广州大学 | Malicious information detection method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118395135A (en) | 2024-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10936947B1 (en) | Recurrent neural network-based artificial intelligence system for time series predictions | |
| Wang et al. | Big data driven cycle time parallel prediction for production planning in wafer manufacturing | |
| US10748072B1 (en) | Intermittent demand forecasting for large inventories | |
| Cadavid et al. | Machine learning in production planning and control: A review of empirical literature | |
| Siregar et al. | Forecasting of raw material needed for plastic products based in income data using ARIMA method | |
| CN117807377B (en) | Multidimensional logistics data mining and predicting method and system | |
| KR102355320B1 (en) | Method for predicting rfid-based material usage using artificial intelligence and providing automatic order system | |
| Turkmen et al. | Intermittent demand forecasting with deep renewal processes | |
| Tan et al. | Analysis of production cycle-time distribution with a big-data approach | |
| CN115689334A (en) | Efficiency analysis method and system of warehouse management system and computer equipment | |
| CN118011990A (en) | Industrial data quality monitoring and improving system based on artificial intelligence | |
| Omri et al. | Data management requirements for phm implementation in smes | |
| KR20210073309A (en) | Demand estimating method and inventory management system using machine learning | |
| CN113742248A (en) | Method and system for predicting organization process based on project measurement data | |
| CN118395135B (en) | Data preprocessing system and method for supply chain planning system | |
| US20130317889A1 (en) | Methods for assessing transition value and devices thereof | |
| CN117151276A (en) | Intelligent management system of electricity selling platform | |
| Bey-Temsamani et al. | A practical approach to combine data mining and prognostics for improved predictive maintenance | |
| Yue et al. | Big data for furniture intelligent manufacturing: conceptual framework, technologies, applications, and challenges | |
| CN115062687A (en) | Enterprise credit monitoring method, device, equipment and storage medium | |
| CN114429297A (en) | Method and device for monitoring risk of project, computer equipment and storage medium | |
| Chouakang et al. | Information technology & supply chain management: role of big data on efficiency | |
| CN118469405B (en) | Processing method, device, equipment and storage medium for household supply chain data | |
| CN118410922B (en) | Data processing method and system based on product supply chain | |
| Tantawy et al. | Applying Big Data Analytics to Retail for Improved Supply Chain Visibility |
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 |