CN117788165B - Enterprise supply chain transaction risk prediction method and system based on artificial intelligence - Google Patents
Enterprise supply chain transaction risk prediction method and system based on artificial intelligence Download PDFInfo
- Publication number
- CN117788165B CN117788165B CN202410212584.2A CN202410212584A CN117788165B CN 117788165 B CN117788165 B CN 117788165B CN 202410212584 A CN202410212584 A CN 202410212584A CN 117788165 B CN117788165 B CN 117788165B
- Authority
- CN
- China
- Prior art keywords
- data
- function
- supply chain
- interval
- historical
- 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
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 31
- 238000011156 evaluation Methods 0.000 claims abstract description 24
- 238000007781 pre-processing Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 14
- 230000008859 change Effects 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 10
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000004140 cleaning Methods 0.000 claims description 9
- 230000006978 adaptation Effects 0.000 claims description 6
- 239000000284 extract Substances 0.000 claims description 6
- 230000001351 cycling effect Effects 0.000 claims description 3
- 238000013480 data collection Methods 0.000 claims description 3
- 230000007423 decrease Effects 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000001303 quality assessment method Methods 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 abstract description 3
- 238000000691 measurement method Methods 0.000 abstract description 3
- 238000005259 measurement Methods 0.000 description 5
- 238000007418 data mining Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000011524 similarity measure Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012502 risk assessment Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an enterprise supply chain transaction risk prediction method and system based on artificial intelligence. The invention belongs to the technical field of image processing, in particular to an enterprise supply chain transaction risk prediction method and system based on artificial intelligence, wherein the scheme constructs a similarity measurement method combining a numerical distance and a curve shape by deducing a specific calculation formula of an interval function Euclidean distance under a basis function and derivative information, reflecting an absolute difference value based on the distance of the basis function, and reflecting the curve shape difference according to the distance of the derivative function information; using the average value of the clustering result evaluation values obtained by taking the data points as initial clustering centers in three adjacent times as a fitness value; the moving strategy considers global optimal individual and random parameters, so that the searching efficiency and the discovery of a local optimal solution are improved.
Description
Technical Field
The invention relates to the technical field of risk prediction, in particular to an enterprise supply chain transaction risk prediction method and system based on artificial intelligence.
Background
Supply chain transaction risk management for modern enterprises is critical to the long-term evolution of the enterprise. With the continuous development and the intelligent progress of the information technology, the artificial intelligence technology is gradually applied to the supply chain transaction risk prediction of enterprises, provides more accurate and reliable risk assessment and prediction information for the enterprises, and helps the enterprises to make more accurate decisions in environments with higher risks. However, in the general enterprise supply chain transaction risk prediction method, data mining is insufficient due to single similarity measurement, similarity measurement focuses on measuring similarity of curves according to numerical distances and ignores change characteristics of curve shapes, and the problem that clustering results are unreasonable when interval value function data are clustered is solved; the problem that the difference between different clusters is low due to improper selection of a common cluster center, the randomness of a method for searching the cluster center is insufficient, and the adaptability of a search strategy is low.
Disclosure of Invention
Aiming at the problems that the data mining is insufficient due to single similarity measurement, the similarity measurement is focused on measuring the similarity of curves according to numerical distances and ignoring the change characteristics of the shapes of the curves, and clustering results are unreasonable when interval value function data are clustered, the method is used for constructing a similarity measurement method combining the numerical distances and the shapes of the curves by deducing a specific calculation formula of the Euclidean distance of an interval function under basis function and derivative information, reflecting the absolute difference of numerical values based on the distance of the basis function, reflecting the shape difference of the curves according to the distance of the derivative function information, and enabling the clustering results to have higher stability, expressive force and accuracy; aiming at the problems of low distinguishability among different clusters caused by improper selection of a general clustering center, insufficient randomness of a method for searching the clustering center and low adaptability of a searching strategy, the method can better capture the stability of the clustering center by using the average value of clustering result evaluation values obtained by taking data points as initial clustering centers for three times adjacently as a fitness value; the global optimal individual and random parameters are considered in the movement strategy, so that the searching efficiency and the discovery of the local optimal solution are improved; thereby improving the searching efficiency and the quality of the clustering result.
The technical scheme adopted by the invention is as follows: the invention provides an artificial intelligence-based enterprise supply chain transaction risk prediction method, which comprises the following steps:
step S1: collecting data;
Step S2: preprocessing data;
step S3: clustering processing based on interval function expansion distance;
step S4: searching an initial cluster center;
Step S5: supply chain transaction risk prediction.
Further, in step S1, the data collection is to collect historical supply chain transaction data, historical supplier assessment data, historical product information, historical market data, historical macro-economic data, and historical supply chain transaction risk;
The historical supply chain transaction data includes order quantity, transaction amount, and transaction date; the historical vendor evaluation data includes a vendor's credit rating, a vendor's historical transaction record, and a vendor's financial status; the historical product information comprises classification of products, quality assessment of the products and prices of the products; the historical market data includes market demand and market competition conditions; the historical macro-economic data includes currency expansion rate, interest rate, and exchange rate.
Further, in step S2, the data preprocessing is to perform data cleaning, data conversion, normalization processing and construction of an interval value function dataset on the collected data; the data cleaning is to perform missing value processing, abnormal value processing and repeated value processing on the collected data; the data conversion is to convert the cleaned data into a vector form based on feature codes; the data normalization is to normalize the converted data based on Min-Max scaling; the construction interval value function data set is to divide the risk of supply chain transaction into three intervals of low, medium and high, and allocate risk level for each interval; and combining the feature vector after the standardization processing with the target variable interval and the risk level to construct an interval value function data set.
Further, in step S3, the clustering process based on the interval function expansion distance specifically includes the following steps:
Step S31: defining an interval function expansion distance, wherein the first half part of the interval function expansion distance extracts information of an interval value function to measure the similarity of a function curve in a numerical distance, and the second half part extracts derivative function information of an upper limit function and a lower limit function to reflect the similarity of the function curve in a curve shape; by combining the two, the numerical distance of the interval value function and the morphological feature are considered, and the numerical information of the interval value function, the similarity between curve shapes and the Euclidean distance of the interval value function are considered; acquiring derivative function information of a lower limit function and an upper limit function of the interval value function; the interval function expansion distance is obtained by combining the original function and the derivative function information, and is expressed as follows:
;
In the method, in the process of the invention, The interval function expansion distance between the ith interval value function sample and the jth interval value function sample at the time point t; t is a time variable; And The lower limit function and the upper limit function of the ith interval value function sample at the time point t are respectively; And The lower limit function and the upper limit function of the jth interval value function sample at the time point t are respectively; And Respectively areAndA first derivative function at a time point t, representing a rate of change of a lower limit function of the interval value function sample; And Respectively areAndA first derivative function at a time point t, representing a rate of change of an interval value function sample upper limit function;
Step S32: establishing an initial cluster centroid, and calculating interval function expansion distances of any two interval value function samples for all interval value function data samples x (t); taking two interval value function samples with the maximum interval function expansion distance as initial clustering centers, and marking as And; Calculating other interval value function samples to the existing clustering centerAndIs the sum of the distances of (2); the interval value function sample with the largest sum of the distances is selected as the third initial clustering center and expressed as; Similarly, all initial cluster centers and cluster numbers are obtained by cycling:;
step S33: data point allocation; calculating the distance from each interval value function sample to the clustering center; distributing data points according to a distance minimum principle; in the mth iteration, the cluster of the ith interval value function samples is represented as:
;
In the method, in the process of the invention, Is the cluster to which the ith interval value function sample belongs in the mth iteration; representing the centroid of cluster s at m-1 iterations; s is an index of clusters, and K is the number of clusters;
step S34: updating a clustering center; taking the data point closest to the average value of all samples in the cluster as a new cluster center; in the mth iteration, the new cluster center may be represented as follows:
;
In the method, in the process of the invention, Is the centroid of the s cluster of the mth iteration; Representing the number of interval value function samples allocated to the s-th cluster;
Step S35: determining a cluster result; repeating steps S33 and S34 until the centroid remains unchanged;
Step S36: and evaluating the clustering result, and considering the similarity of samples in the classes and the difference of samples between the classes, wherein the similarity and the difference are expressed as follows:
;
Wherein PG is a cluster evaluation index; n is the number of categories; And The average interval function expansion distance of the data points in the ith category and the jth category is respectively; the interval function expansion distance sum of the data points in the ith class and the jth class;
step S37: and (3) assigning cluster labels, wherein the supply chain transaction risk level is used as a label dimension, and the supply chain transaction risk level with the maximum number of sample data is used as the cluster label.
Further, in step S4, the initial cluster center search is that an fitness threshold is preset, and when the cluster evaluation index obtained in step S3 is higher than the fitness threshold, a search strategy is adopted to search the initial cluster center, which specifically includes the following steps:
Step S41: initializing a clustering center, determining a parameter searching space based on a data sample space, randomly selecting a group of sample data points as initial clustering center points, wherein the fitness value of each data point is the average value of clustering result evaluation values obtained by taking the data point as the initial clustering center three times adjacently; taking 10% of data points with the lowest fitness value as a random group, and the rest data points as search groups; when the mobile device moves, a nearby strategy is adopted, and the searched data points are the data points closest to the position after the mobile device moves;
Step S42: defining a random group movement strategy, wherein random group individuals pay attention to the randomness of search, and only global optimal individuals and random parameters are considered during movement, and the following formula is adopted:
;
;
Wherein r is a random parameter, which decreases with increasing iteration number; d is the dimension size; t is the current iteration number, and T max is the maximum iteration number; And The individual positions of the I random group at the T+1st iteration and the T iteration are respectively; is a globally optimal individual location; rand (0, 1) is a random number in the range of 0 to 1;
Step S43: defining a search group movement strategy, and updating the search group individuals according to the historical positions and the population optimal positions of the search group individuals by using the following formula:
;
;
Wherein X rand is a random individual; And The individual positions of the I search group are respectively the T+1st iteration and the T iteration; f (·) is an fitness value function; is the optimal position of the history of the individual; θ is a random number between 0 and 360;
Step S44: searching and judging, and outputting a searched initial cluster center if the adaptation value of the initial cluster center searched currently is lower than the adaptation threshold; if the maximum iteration number is reached, re-initializing the clustering center and searching; otherwise, continuing to move the position search.
Further, in step S5, the supply chain transaction risk prediction is to collect supply chain transaction data, supplier assessment data, product information, market data and macro economic data in real time as real time data; taking the data acquired in the step S1 as historical data; and the real-time data and the historical data participate in clustering together, and cluster labels corresponding to the clustered real-time data are used as enterprise supply chain transaction risk prediction results.
The invention provides an artificial intelligence-based enterprise supply chain transaction risk prediction system, which comprises a data acquisition module, a data preprocessing module, a clustering processing module, an initial clustering center searching module and a supply chain transaction risk prediction module, wherein the data acquisition module is used for acquiring data of an enterprise;
the data acquisition module acquires historical supply chain transaction data, historical provider evaluation data, historical product information, historical market data, historical macro-economic data and historical supply chain transaction risk, and sends the data to the data preprocessing module;
The data preprocessing module performs data cleaning, data conversion, standardization processing and construction of interval value function data sets on the acquired data, and sends the data to the clustering processing module;
the clustering processing module is used for defining an interval function expansion distance by combining the original function and derivative function information, and completing clustering through initial cluster centroid establishment, data point distribution and cluster center updating; transmitting the data to an initial cluster center searching module;
the initial cluster center searching module is used for defining a random group movement strategy and a search group movement strategy to realize cluster center searching based on the initial cluster center and the defined fitness value, and sending data to the supply chain transaction risk prediction module;
The supply chain transaction risk prediction module collects supply chain transaction data, supplier evaluation data, product information, market data and macro economic data in real time as real-time data, and achieves supply chain transaction risk prediction based on clustering results.
By adopting the scheme, the beneficial effects obtained by the invention are as follows:
(1) Aiming at the problem that the data mining is insufficient due to the fact that a single similarity measure exists in a general enterprise supply chain transaction risk prediction method, the similarity measurement is focused on the problem that the change characteristics of curve shapes are ignored when the similarity of curves is measured according to the numerical distance, and clustering results are unreasonable when interval value function data are clustered, the method is characterized in that the absolute difference value of numerical values is reflected based on the distance of a base function and a specific calculation formula of Euclidean distance of an interval function under derivative information, the similarity measurement method combining the numerical distance and the curve shapes is constructed according to the distance of the derivative function information, and therefore the clustering results are higher in stability, expressive force and accuracy.
(2) Aiming at the problems of low distinguishability among different clusters caused by improper selection of a general clustering center, insufficient randomness of a method for searching the clustering center and low adaptability of a searching strategy, the method can better capture the stability of the clustering center by using the average value of clustering result evaluation values obtained by taking data points as initial clustering centers for three times adjacently as a fitness value; the global optimal individual and random parameters are considered in the movement strategy, so that the searching efficiency and the discovery of the local optimal solution are improved; thereby improving the searching efficiency and the quality of the clustering result.
Drawings
FIG. 1 is a schematic flow chart of an artificial intelligence based enterprise supply chain transaction risk prediction method provided by the invention;
FIG. 2 is a schematic diagram of an artificial intelligence based enterprise supply chain transaction risk prediction system according to the present invention;
FIG. 3 is a flow chart of step S3;
fig. 4 is a flow chart of step S4.
The accompanying drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate the invention and together with the embodiments of the invention, serve to explain the invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention; all other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In the description of the present invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "top," "bottom," "inner," "outer," and the like indicate orientation or positional relationships based on those shown in the drawings, merely to facilitate describing the present invention and simplify the description, and do not indicate or imply that the system or element being referred to must have a particular orientation, be constructed and operate in a particular orientation, and thus should not be construed as limiting the present invention.
Referring to fig. 1, the method for predicting risk of enterprise supply chain transaction based on artificial intelligence provided by the invention comprises the following steps:
Step S1: data collection, namely collecting historical supply chain transaction data, historical provider evaluation data, historical product information, historical market data, historical macro-economic data and historical supply chain transaction risk;
Step S2: data preprocessing, namely performing data cleaning, data conversion, standardization processing and construction of interval value function data sets on the acquired data;
step S3: based on the clustering processing of the interval function expansion distance, the interval function expansion distance is defined by combining the original function and derivative function information, and clustering is completed through initial cluster centroid establishment, data point distribution and cluster center updating;
Step S4: the method comprises the steps of initial cluster center searching, defining a random group movement strategy and a search group movement strategy to realize cluster center searching based on an initial cluster center and a defined fitness value;
Step S5: supply chain transaction risk prediction.
In a second embodiment, referring to fig. 1, the embodiment is based on the above embodiment, and in step S1, the historical supply chain transaction data includes an order number, a transaction amount, and a transaction date; the historical vendor evaluation data includes a vendor's credit rating, a vendor's historical transaction record, and a vendor's financial status; the historical product information comprises classification of products, quality assessment of the products and prices of the products; the historical market data includes market demand and market competition conditions; the historical macro-economic data includes currency expansion rate, interest rate, and exchange rate.
An embodiment III, referring to FIG. 1, based on the above embodiment, the data cleaning in step S2 is to perform missing value processing, outlier processing and repeated value processing on the collected data; the data conversion is to convert the cleaned data into a vector form based on feature codes; the data normalization is to normalize the converted data based on Min-Max scaling; the construction interval value function data set is to divide the risk of supply chain transaction into three intervals of low, medium and high, and allocate risk level for each interval; and combining the feature vector after the standardization processing with the target variable interval and the risk level to construct an interval value function data set.
In the fourth embodiment, referring to fig. 1 and 3, the clustering process based on the interval function expansion distance in step S3 specifically includes the following steps:
Step S31: defining an interval function expansion distance, wherein the first half part of the interval function expansion distance extracts information of an interval value function to measure the similarity of a function curve in a numerical distance, and the second half part extracts derivative function information of an upper limit function and a lower limit function to reflect the similarity of the function curve in a curve shape; by combining the two, the numerical distance of the interval value function and the morphological feature are considered, and the numerical information of the interval value function, the similarity between curve shapes and the Euclidean distance of the interval value function are considered; acquiring derivative function information of a lower limit function and an upper limit function of the interval value function; the interval function expansion distance is obtained by combining the original function and the derivative function information, and is expressed as follows:
;
In the method, in the process of the invention, The interval function expansion distance between the ith interval value function sample and the jth interval value function sample at the time point t; t is a time variable; And The lower limit function and the upper limit function of the ith interval value function sample at the time point t are respectively; And The lower limit function and the upper limit function of the jth interval value function sample at the time point t are respectively; And Respectively areAndA first derivative function at a time point t, representing a rate of change of a lower limit function of the interval value function sample; And Respectively areAndA first derivative function at a time point t, representing a rate of change of an interval value function sample upper limit function;
Step S32: establishing an initial cluster centroid, and calculating interval function expansion distances of any two interval value function samples for all interval value function data samples x (t); taking two interval value function samples with the maximum interval function expansion distance as initial clustering centers, and marking as And; Calculating other interval value function samples to the existing clustering centerAndIs the sum of the distances of (2); the interval value function sample with the largest sum of the distances is selected as the third initial clustering center and expressed as; Similarly, all initial cluster centers and cluster numbers are obtained by cycling:;
step S33: data point allocation; calculating the distance from each interval value function sample to the clustering center; distributing data points according to a distance minimum principle; in the mth iteration, the cluster of the ith interval value function samples is represented as:
;
In the method, in the process of the invention, Is the cluster to which the ith interval value function sample belongs in the mth iteration; representing the centroid of cluster s at m-1 iterations; s is an index of clusters, and K is the number of clusters;
step S34: updating a clustering center; taking the data point closest to the average value of all samples in the cluster as a new cluster center; in the mth iteration, the new cluster center may be represented as follows:
;
In the method, in the process of the invention, Is the centroid of the s cluster of the mth iteration; Representing the number of interval value function samples allocated to the s-th cluster;
Step S35: determining a cluster result; repeating steps S33 and S34 until the centroid remains unchanged;
Step S36: and evaluating the clustering result, and considering the similarity of samples in the classes and the difference of samples between the classes, wherein the similarity and the difference are expressed as follows:
;
Wherein PG is a cluster evaluation index; n is the number of categories; And The average interval function expansion distance of the data points in the ith category and the jth category is respectively; the interval function expansion distance sum of the data points in the ith class and the jth class;
step S37: and (3) assigning cluster labels, wherein the supply chain transaction risk level is used as a label dimension, and the supply chain transaction risk level with the maximum number of sample data is used as the cluster label.
By executing the above operation, the method aims at the problem that the data mining is insufficient due to the fact that a single similarity measure exists in a general enterprise supply chain transaction risk prediction method, the similarity measure focuses on measuring the similarity of curves according to the numerical distance and ignoring the change characteristics of the shapes of the curves, and clustering results are unreasonable when interval value function data are clustered.
In step S4, an fitness threshold is preset for the initial cluster center search, and when the cluster evaluation index obtained in step S3 is higher than the fitness threshold, a search strategy is adopted to search the initial cluster center, which specifically includes the following steps:
Step S41: initializing a clustering center, determining a parameter searching space based on a data sample space, randomly selecting a group of sample data points as initial clustering center points, wherein the fitness value of each data point is the average value of clustering result evaluation values obtained by taking the data point as the initial clustering center three times adjacently; taking 10% of data points with the lowest fitness value as a random group, and the rest data points as search groups; when the mobile device moves, a nearby strategy is adopted, and the searched data points are the data points closest to the position after the mobile device moves;
Step S42: defining a random group movement strategy, wherein random group individuals pay attention to the randomness of search, and only global optimal individuals and random parameters are considered during movement, and the following formula is adopted:
;
;
Wherein r is a random parameter, which decreases with increasing iteration number; d is the dimension size; t is the current iteration number, and T max is the maximum iteration number; And The individual positions of the I random group at the T+1st iteration and the T iteration are respectively; is a globally optimal individual location; rand (0, 1) is a random number in the range of 0 to 1;
Step S43: defining a search group movement strategy, and updating the search group individuals according to the historical positions and the population optimal positions of the search group individuals by using the following formula:
;
;
Wherein X rand is a random individual; And The individual positions of the I search group are respectively the T+1st iteration and the T iteration; f (·) is an fitness value function; is the optimal position of the history of the individual; θ is a random number between 0 and 360;
Step S44: searching and judging, and outputting a searched initial cluster center if the adaptation value of the initial cluster center searched currently is lower than the adaptation threshold; if the maximum iteration number is reached, re-initializing the clustering center and searching; otherwise, continuing to move the position search.
By executing the operation, aiming at the problems of low distinguishability among different clusters caused by improper selection of a general clustering center, insufficient randomness of a method for searching the clustering center and low adaptability of a searching strategy, the method can better capture the stability of the clustering center by using the average value of clustering result evaluation values obtained by taking data points as initial clustering centers for three adjacent times as a fitness value; the global optimal individual and random parameters are considered in the movement strategy, so that the searching efficiency and the discovery of the local optimal solution are improved; thereby improving the searching efficiency and the quality of the clustering result.
Embodiment six, referring to fig. 1, based on the above embodiment, in step S5, the supply chain transaction risk prediction is to collect supply chain transaction data, supplier evaluation data, product information, market data, and macro economic data as real-time data; taking the data acquired in the step S1 as historical data; and the real-time data and the historical data participate in clustering together, and cluster labels corresponding to the clustered real-time data are used as enterprise supply chain transaction risk prediction results.
An embodiment seven, referring to fig. 2, based on the above embodiment, the enterprise supply chain transaction risk prediction system based on artificial intelligence provided by the invention includes a data acquisition module, a data preprocessing module, a clustering processing module, an initial clustering center searching module and a supply chain transaction risk prediction module;
the data acquisition module acquires historical supply chain transaction data, historical provider evaluation data, historical product information, historical market data, historical macro-economic data and historical supply chain transaction risk, and sends the data to the data preprocessing module;
The data preprocessing module performs data cleaning, data conversion, standardization processing and construction of interval value function data sets on the acquired data, and sends the data to the clustering processing module;
the clustering processing module is used for defining an interval function expansion distance by combining the original function and derivative function information, and completing clustering through initial cluster centroid establishment, data point distribution and cluster center updating; transmitting the data to an initial cluster center searching module;
the initial cluster center searching module is used for defining a random group movement strategy and a search group movement strategy to realize cluster center searching based on the initial cluster center and the defined fitness value, and sending data to the supply chain transaction risk prediction module;
The supply chain transaction risk prediction module collects supply chain transaction data, supplier evaluation data, product information, market data and macro economic data in real time as real-time data, and achieves supply chain transaction risk prediction based on clustering results.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made hereto without departing from the spirit and principles of the present invention.
The invention and its embodiments have been described above with no limitation, and the actual construction is not limited to the embodiments of the invention as shown in the drawings. In summary, if one of ordinary skill in the art is informed by this disclosure, a structural manner and an embodiment similar to the technical solution should not be creatively devised without departing from the gist of the present invention.
Claims (2)
1. The enterprise supply chain transaction risk prediction method based on artificial intelligence is characterized by comprising the following steps of: the method comprises the following steps:
step S1: collecting data;
Step S2: preprocessing data;
step S3: clustering processing based on interval function expansion distance;
step S4: searching an initial cluster center;
Step S5: supply chain transaction risk prediction;
In step S3, the clustering process based on the interval function expansion distance specifically includes the following steps:
Step S31: defining an interval function expansion distance, wherein the first half part of the interval function expansion distance extracts information of an interval value function to measure the similarity of a function curve in a numerical distance, and the second half part extracts derivative function information of an upper limit function and a lower limit function to reflect the similarity of the function curve in a curve shape; by combining the two, the numerical distance of the interval value function and the morphological feature are considered, and the numerical information of the interval value function, the similarity between curve shapes and the Euclidean distance of the interval value function are considered; acquiring derivative function information of a lower limit function and an upper limit function of the interval value function; the interval function expansion distance is obtained by combining the original function and the derivative function information, and is expressed as follows:
;
In the method, in the process of the invention, The interval function expansion distance between the ith interval value function sample and the jth interval value function sample at the time point t; t is a time variable; /(I)And/>The lower limit function and the upper limit function of the ith interval value function sample at the time point t are respectively; /(I)And/>The lower limit function and the upper limit function of the jth interval value function sample at the time point t are respectively; /(I)And/>Are respectively/>And/>A first derivative function at a time point t, representing a rate of change of a lower limit function of the interval value function sample; /(I)And/>Are respectively/>And/>A first derivative function at a time point t, representing a rate of change of an interval value function sample upper limit function;
Step S32: establishing an initial cluster centroid, and calculating interval function expansion distances of any two interval value function samples for all interval value function data samples x (t); taking two interval value function samples with the maximum interval function expansion distance as initial clustering centers, and marking as And/>; Calculating other interval value function samples to the existing clustering center/>And/>Is the sum of the distances of (2); the interval value function sample with the largest sum of the distances is selected as the third initial cluster center and expressed as/>; Similarly, all initial cluster centers and cluster numbers are obtained by cycling: /(I);
Step S33: data point allocation; calculating the distance from each interval value function sample to the clustering center; distributing data points according to a distance minimum principle; in the mth iteration, the cluster of the ith interval value function samples is represented as:
;
In the method, in the process of the invention, Is the cluster to which the ith interval value function sample belongs in the mth iteration; /(I)Representing the centroid of cluster s at m-1 iterations; s is an index of clusters, and K is the number of clusters;
step S34: updating a clustering center; taking the data point closest to the average value of all samples in the cluster as a new cluster center; in the mth iteration, the new cluster center may be represented as follows:
;
In the method, in the process of the invention, Is the centroid of the s cluster of the mth iteration; /(I)Representing the number of interval value function samples allocated to the s-th cluster;
Step S35: determining a cluster result; repeating steps S33 and S34 until the centroid remains unchanged;
Step S36: and evaluating the clustering result, and considering the similarity of samples in the classes and the difference of samples between the classes, wherein the similarity and the difference are expressed as follows:
;
Wherein PG is a cluster evaluation index; n is the number of categories; and/> The average interval function expansion distance of the data points in the ith category and the jth category is respectively; /(I)The interval function expansion distance sum of the data points in the ith class and the jth class;
Step S37: assigning cluster labels, wherein the supply chain transaction risk level is used as a label dimension, and the supply chain transaction risk level with the maximum amount of sample data is used as a cluster label;
In step S4, the initial cluster center search is preset with an fitness threshold, and when the cluster evaluation index obtained in step S3 is higher than the fitness threshold, a search strategy is adopted to search the initial cluster center, which specifically includes the following steps:
Step S41: initializing a clustering center, determining a parameter searching space based on a data sample space, randomly selecting a group of sample data points as initial clustering center points, wherein the fitness value of each data point is the average value of clustering result evaluation values obtained by taking the data point as the initial clustering center three times adjacently; taking 10% of data points with the lowest fitness value as a random group, and the rest data points as search groups; when the mobile device moves, a nearby strategy is adopted, and the searched data points are the data points closest to the position after the mobile device moves;
Step S42: defining a random group movement strategy, wherein random group individuals pay attention to the randomness of search, and only global optimal individuals and random parameters are considered during movement, and the following formula is adopted:
;
;
Wherein r is a random parameter, which decreases with increasing iteration number; d is the dimension size; t is the current iteration number, and T max is the maximum iteration number; and/> The individual positions of the I random group at the T+1st iteration and the T iteration are respectively; /(I)Is a globally optimal individual location; rand (0, 1) is a random number in the range of 0 to 1;
Step S43: defining a search group movement strategy, and updating the search group individuals according to the historical positions and the population optimal positions of the search group individuals by using the following formula:
;
;
Wherein X rand is a random individual; and/> The individual positions of the I search group are respectively the T+1st iteration and the T iteration; f (·) is an fitness value function; /(I)Is the optimal position of the history of the individual; θ is a random number between 0 and 360;
step S44: searching and judging, and outputting a searched initial cluster center if the adaptation value of the initial cluster center searched currently is lower than the adaptation threshold; if the maximum iteration number is reached, re-initializing the clustering center and searching; otherwise, continuing to search the mobile position;
In step S1, the data collection is to collect historical supply chain transaction data, historical vendor assessment data, historical product information, historical market data, historical macro-economic data, and historical supply chain transaction risk;
The historical supply chain transaction data includes order quantity, transaction amount, and transaction date; the historical vendor evaluation data includes a vendor's credit rating, a vendor's historical transaction record, and a vendor's financial status; the historical product information comprises classification of products, quality assessment of the products and prices of the products; the historical market data includes market demand and market competition conditions; the historical macro-economic data includes a currency expansion rate, an interest rate, and an exchange rate;
In step S2, the data preprocessing is to perform data cleaning, data conversion, standardization processing and construction of an interval value function data set on the collected data; the data cleaning is to perform missing value processing, abnormal value processing and repeated value processing on the collected data; the data conversion is to convert the cleaned data into a vector form based on feature codes; the data standardization is to standardize the converted data based on Min-Max scaling; the construction interval value function data set is to divide the risk of supply chain transaction into three intervals of low, medium and high, and allocate risk level for each interval; combining the feature vector after the standardization processing with the target variable interval and the risk level to construct an interval value function data set;
In step S5, the supply chain transaction risk prediction is to collect supply chain transaction data, supplier assessment data, product information, market data and macro economic data in real time as real-time data; taking the data acquired in the step S1 as historical data; and the real-time data and the historical data participate in clustering together, and cluster labels corresponding to the clustered real-time data are used as enterprise supply chain transaction risk prediction results.
2. An artificial intelligence based enterprise supply chain transaction risk prediction system for implementing the artificial intelligence based enterprise supply chain transaction risk prediction method as claimed in claim 1, wherein: the system comprises a data acquisition module, a data preprocessing module, a clustering processing module, an initial clustering center searching module and a supply chain transaction risk prediction module;
the data acquisition module acquires historical supply chain transaction data, historical provider evaluation data, historical product information, historical market data, historical macro-economic data and historical supply chain transaction risk, and sends the data to the data preprocessing module;
The data preprocessing module performs data cleaning, data conversion, standardization processing and construction of interval value function data sets on the acquired data, and sends the data to the clustering processing module;
the clustering processing module is used for defining an interval function expansion distance by combining the original function and derivative function information, and completing clustering through initial cluster centroid establishment, data point distribution and cluster center updating; transmitting the data to an initial cluster center searching module;
the initial cluster center searching module is used for defining a random group movement strategy and a search group movement strategy to realize cluster center searching based on the initial cluster center and the defined fitness value, and sending data to the supply chain transaction risk prediction module;
The supply chain transaction risk prediction module collects supply chain transaction data, supplier evaluation data, product information, market data and macro economic data in real time as real-time data, and achieves supply chain transaction risk prediction based on clustering results.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410212584.2A CN117788165B (en) | 2024-02-27 | 2024-02-27 | Enterprise supply chain transaction risk prediction method and system based on artificial intelligence |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410212584.2A CN117788165B (en) | 2024-02-27 | 2024-02-27 | Enterprise supply chain transaction risk prediction method and system based on artificial intelligence |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117788165A CN117788165A (en) | 2024-03-29 |
CN117788165B true CN117788165B (en) | 2024-05-14 |
Family
ID=90387275
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410212584.2A Active CN117788165B (en) | 2024-02-27 | 2024-02-27 | Enterprise supply chain transaction risk prediction method and system based on artificial intelligence |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117788165B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN118228075A (en) * | 2024-05-23 | 2024-06-21 | 国网山东省电力公司宁津县供电公司 | Single-user power failure automatic alarm monitoring method and system based on big data |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE19912126A1 (en) * | 1999-03-18 | 2000-09-21 | Abb Research Ltd | Reconstruction of time value data series from plotter charts printed on paper rolls by reading the curves in with a scanner and converting them to digitized data using mathematical functions for separating different colors |
EP1193648A1 (en) * | 2000-06-30 | 2002-04-03 | Texas Instruments Incorporated | Image object ranking |
CN109242016A (en) * | 2018-08-30 | 2019-01-18 | 天津理工大学 | A kind of similitude judgment method of space curve |
CN109902953A (en) * | 2019-02-27 | 2019-06-18 | 华北电力大学 | A kind of classification of power customers method based on adaptive population cluster |
CN113935247A (en) * | 2021-11-05 | 2022-01-14 | 国网甘肃省电力公司电力科学研究院 | Partition virtual inertia estimation method considering wind speed randomness and correlation |
CN115051346A (en) * | 2022-01-05 | 2022-09-13 | 国网浙江省电力有限公司常山县供电公司 | Distributed photovoltaic output prediction method |
CN115809719A (en) * | 2022-10-17 | 2023-03-17 | 国网上海市电力公司 | Short-term load prediction correction method based on morphological clustering |
WO2023201552A1 (en) * | 2022-04-20 | 2023-10-26 | Southeast University | County-wide photovoltaic prediction method based on cluster division and data enhancement |
CN116976574A (en) * | 2022-04-18 | 2023-10-31 | 天津大学 | Building load curve dimension reduction method based on two-stage hybrid clustering algorithm |
CN117114144A (en) * | 2023-10-24 | 2023-11-24 | 青岛农业大学 | Rice salt and alkali resistance prediction method and system based on artificial intelligence |
CN117273237A (en) * | 2023-11-15 | 2023-12-22 | 宁德市天铭新能源汽车配件有限公司 | Method and system for predicting capacity of automobile parts based on artificial intelligence |
-
2024
- 2024-02-27 CN CN202410212584.2A patent/CN117788165B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE19912126A1 (en) * | 1999-03-18 | 2000-09-21 | Abb Research Ltd | Reconstruction of time value data series from plotter charts printed on paper rolls by reading the curves in with a scanner and converting them to digitized data using mathematical functions for separating different colors |
EP1193648A1 (en) * | 2000-06-30 | 2002-04-03 | Texas Instruments Incorporated | Image object ranking |
CN109242016A (en) * | 2018-08-30 | 2019-01-18 | 天津理工大学 | A kind of similitude judgment method of space curve |
CN109902953A (en) * | 2019-02-27 | 2019-06-18 | 华北电力大学 | A kind of classification of power customers method based on adaptive population cluster |
CN113935247A (en) * | 2021-11-05 | 2022-01-14 | 国网甘肃省电力公司电力科学研究院 | Partition virtual inertia estimation method considering wind speed randomness and correlation |
CN115051346A (en) * | 2022-01-05 | 2022-09-13 | 国网浙江省电力有限公司常山县供电公司 | Distributed photovoltaic output prediction method |
CN116976574A (en) * | 2022-04-18 | 2023-10-31 | 天津大学 | Building load curve dimension reduction method based on two-stage hybrid clustering algorithm |
WO2023201552A1 (en) * | 2022-04-20 | 2023-10-26 | Southeast University | County-wide photovoltaic prediction method based on cluster division and data enhancement |
CN115809719A (en) * | 2022-10-17 | 2023-03-17 | 国网上海市电力公司 | Short-term load prediction correction method based on morphological clustering |
CN117114144A (en) * | 2023-10-24 | 2023-11-24 | 青岛农业大学 | Rice salt and alkali resistance prediction method and system based on artificial intelligence |
CN117273237A (en) * | 2023-11-15 | 2023-12-22 | 宁德市天铭新能源汽车配件有限公司 | Method and system for predicting capacity of automobile parts based on artificial intelligence |
Non-Patent Citations (1)
Title |
---|
目标轮廓直线特征的高精度提取;李彩林;郭宝云;李畅;;测绘学报;20110215(第01期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN117788165A (en) | 2024-03-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN117788165B (en) | Enterprise supply chain transaction risk prediction method and system based on artificial intelligence | |
Rajabi et al. | A comparative study of clustering techniques for electrical load pattern segmentation | |
Gerbec et al. | Allocation of the load profiles to consumers using probabilistic neural networks | |
Jin et al. | Comparison of clustering techniques for residential energy behavior using smart meter data | |
CN112561156A (en) | Short-term power load prediction method based on user load mode classification | |
CN111324642A (en) | Model algorithm type selection and evaluation method for power grid big data analysis | |
CN105260815A (en) | Electric power system payment channel evaluating method based on large data | |
CN107248086A (en) | Advertisement putting aided analysis method based on user power utilization behavioural analysis | |
CN117670066B (en) | Questor management method, system, equipment and storage medium based on intelligent decision | |
CN112819299A (en) | Differential K-means load clustering method based on center optimization | |
Aeron et al. | Data mining framework for customer lifetime value-based segmentation | |
CN114997841B (en) | Low-carbon behavior data management system based on block chain | |
Razavi et al. | Rethinking the privacy of the smart grid: What your smart meter data can reveal about your household in Ireland | |
CN116976529A (en) | Cross-river-basin water diversion method and system based on supply-demand prediction dynamic correction | |
Tripathy et al. | Forecasting of renewable generation for applications in smart grid power systems | |
Arsene et al. | Profiling consumers in a water distribution network using K-Means clustering and multiple pre-processing methods | |
Fusco et al. | Bayesian network clustering and self-organizing maps under the test of Indian districts. A comparison | |
CN117291655A (en) | Consumer life cycle operation analysis method based on entity and network collaborative mapping | |
CN117217665A (en) | Intelligent inventory management and replenishment strategy generation system | |
CN111798152A (en) | Intelligent store management method and device | |
Tang et al. | Leveraging socioeconomic information and deep learning for residential load pattern prediction | |
Toussaint | Evaluation of clustering techniques for generating household energy consumption patterns in a developing country | |
Kumar et al. | A deep clustering framework for load pattern segmentation | |
CN113191570A (en) | Fund planning recommendation method, device and equipment based on deep learning | |
CN117541200B (en) | Project management method and system based on LTC flow |
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 |