CN113222707A - Intelligent service transaction recommendation method and system - Google Patents
Intelligent service transaction recommendation method and system Download PDFInfo
- Publication number
- CN113222707A CN113222707A CN202110571801.3A CN202110571801A CN113222707A CN 113222707 A CN113222707 A CN 113222707A CN 202110571801 A CN202110571801 A CN 202110571801A CN 113222707 A CN113222707 A CN 113222707A
- Authority
- CN
- China
- Prior art keywords
- model
- recommendation
- data
- vector
- transaction
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 230000000694 effects Effects 0.000 claims abstract description 19
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims abstract description 11
- 230000003993 interaction Effects 0.000 claims abstract description 10
- 238000013528 artificial neural network Methods 0.000 claims abstract description 8
- 239000013598 vector Substances 0.000 claims description 55
- 238000012549 training Methods 0.000 claims description 10
- 238000012795 verification Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012552 review Methods 0.000 description 2
- 241001494479 Pecora Species 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003014 reinforcing effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Abstract
The invention belongs to the technical field of artificial intelligence, and relates to an intelligent service transaction recommendation method and system, which comprises the following steps: s1, acquiring the preprocessed data; s2, constructing a neural network and a learnable matrix decomposition model, inputting the data obtained in the step S1 to train the model, and obtaining a recommendation model with the best effect; s3, inputting the actual demand into the recommendation model with the best effect, and obtaining the service transaction recommendation. The method and the system model many-to-many characteristic interaction between the user and the product, capture more transaction information, improve the recommendation accuracy and better mine some potential requirements of the user.
Description
Technical Field
The invention relates to an intelligent service transaction recommendation method and system, belongs to the technical field of artificial intelligence, and particularly relates to the technical field of intelligent service recommendation.
Background
In recent years, with the continuous development of electronic commerce technology, the number of network products and consumers has increased explosively, and both consumers and merchants face huge amounts of information. It is also a complicated task for a consumer to select a suitable product from a plurality of products, browse, review information of the product, and find a product suitable for his/her own needs. The same problem exists in the transaction between enterprises, and the enterprises often have the need of purchasing a certain commodity, on one hand, the enterprise needs to select a proper commodity from a plurality of enterprise-level commodities when purchasing the commodity, which requires that the enterprise has enough knowledge about the commodity and the enterprise when purchasing the commodity; on the other hand, enterprises may have potential needs for certain commodities, which are often discovered under the condition that the enterprises have caused certain losses, namely, measures for reinforcing the sheep death, and the needs are usually dealt with in time, so that it is difficult to have sufficient time to select the best remedy. Such as the network security facilities of an internet company, etc. In summary, if a better recommendation system exists, matching relationships between consumers and commodities or between enterprises and commodities can be mined according to existing transaction data, interaction between purchasers and commodity features can be found, non-occurring transactions can be predicted, transaction events which are more likely to occur under specific situations can be given, workload of commodity review of purchasers can be reduced, potential needs of enterprises can be found in advance, and losses caused by the fact that the potential needs cannot be met are avoided.
With the vigorous development of modern service industry in China, service transactions have occurred in various scenes of life. The service transaction has the characteristics of diversified transaction modes, massive transaction data, high frequency of transaction requests and the like, so that higher requirements are put forward on the intelligent recommendation system.
The conventional recommendation model is applied to an intelligent service transaction scene at present, and the following problems still exist:
1) the traditional recommendation model only models a one-to-one feature interaction process between a consumer and a commodity, namely modeling the interaction generated by a certain feature of the consumer and a certain feature of the commodity. In fact, many-to-many feature interaction exists, and the accuracy of the recommendation system is reduced due to the fact that partial information is lost.
2) The traditional recommendation algorithm needs to perform complicated processing on the characteristics of consumers and commodities, so that the recommendation system is single in use scene, and when the scene is changed, the method needs to be adjusted in a large scale, and special requirements of diversification of transaction modes, large amount of transaction data, high frequency of transaction requests and the like in a service transaction scene cannot be met.
Disclosure of Invention
In view of the above problems, an object of the present invention is to provide an intelligent service transaction recommendation method, system, readable medium and device, which model many-to-many feature interactions between users and products, capture more transaction information, improve recommendation accuracy, and better mine some potential requirements of users.
In order to achieve the purpose, the invention adopts the following technical scheme: an intelligent service transaction recommendation method comprises the following steps: s1, acquiring the preprocessed data; s2, constructing a neural network and a learnable matrix decomposition model, inputting the data obtained in the step S1 to train the model, and obtaining a recommendation model with the best effect; s3, inputting the actual demand into the recommendation model with the best effect, and obtaining the service transaction recommendation.
Further, the data in step S1 includes user data, service product data, and transaction data, and the user data, service product data, and transaction data obtained in each transaction are concatenated to generate a new vector.
Further, the method for generating the new vector by splicing comprises the following steps: for information of digital type, directly converting into vector; uniformly converting the information of the time types into days; for the text type information, the text is converted into word bags by word segmentation, and then the text vectorization is carried out.
Further, the preprocessing method of the data in step S1 is: and dividing the spliced new vectors into N domains, setting definition targets of products corresponding to the new vectors in all transactions, and dividing the new vectors in all transactions into a training set and a verification set.
Further, if the new vector corresponds to the product, the definition target y of the new vector corresponds to 1, otherwise, the definition target y of the new vector corresponds to 0.
Further, the method for training the model in step S2 is as follows: s2.1 defining an objective function
The method is used for predicting the probability of one transaction generated at a certain moment and determining the size K of a decomposition matrix; s2.2 two fully-connected networks are constructed for each domain of the new vector, and A is calculated according to output values of the two fully-connected networks1(ii) a S2.3, inputting the output vector M of the second one of the two fully-connected networks into the model, wherein the output result of the model is A2(ii) a S2.4 prediction results
According to the prediction result
And defining an object y, and updating the weight of the fully-connected network until a best-effect recommendation model is obtained.
Further, after the step S2.4 is finished, inputting the data in the verification set into the recommended model with the best effect, if the verification result is improved, saving the model, otherwise, repeating the steps S2.1-S2.4, and retraining the model.
Further, in step S2.2A1The calculation formula of (2) is as follows:
where N is the number of fields, oiIs the output value of the first fully connected network; m isiAnd mjAre output values of the second fully connected network.
Further, step S3 specifically includes the following steps: s3.1, acquiring an actual demand vector and a transaction data vector; s3.2 input vector [ p ] of product numbered ii1，pi2,…,piy]Generating a vector xiIf the total number of products is n, generating all product vector sets Z ═ Z1,z2,…，zn}；S3.3 inputting all product vector sets Z into the best recommendation model and outputting a prediction result set
And S3.4, sorting the prediction results in the prediction result set from big to small, and recommending the results ranked in the front.
The invention also discloses an intelligent service transaction recommendation system, which comprises: the data acquisition module is used for acquiring the preprocessed data; the model training module is used for constructing a neural network and a learnable matrix decomposition model, inputting the data obtained in the data obtaining module to train the model and obtaining a recommendation model with the best effect; and the interaction module is used for inputting the actual requirements into the recommendation model with the best effect to obtain service transaction recommendation.
Due to the adoption of the technical scheme, the invention has the following advantages:
1. the deep neural network and learnable matrix decomposition-based recommendation system is adopted to model many-to-many characteristic interaction between users and products, capture more transaction information, improve the recommendation accuracy and better mine some potential requirements of the users.
2. The invention can carry out universal processing on the characteristics of the user and the commodity, does not need to carry out large-scale adjustment on the recommendation method when the scene is changed, and can meet the requirements of service transaction scenes.
Drawings
Fig. 1 is a schematic diagram of a method for recommending smart service transactions in an embodiment of the present invention.
Detailed Description
The present invention is described in detail by way of specific embodiments in order to better understand the technical direction of the present invention for those skilled in the art. It should be understood, however, that the detailed description is provided for a better understanding of the invention only and that they should not be taken as limiting the invention. In describing the present invention, it is to be understood that the terminology used is for the purpose of description only and is not intended to be indicative or implied of relative importance.
Example one
The intelligent service transaction recommendation method of the embodiment, as shown in fig. 1, includes the following steps:
and S1, acquiring the preprocessed data, reading and connecting the data set, and providing data for model training.
The preprocessed data set is a transaction data set of the safety service product, the data in the data set comprises user data, service product data and transaction data, and each user data is generated into a vector [ c ]1,c2,...,cx](ii) a Generating service product data into vector p1,p2,…,py](ii) a Generating transaction data into vector t1,t2,…,tz]。
Splicing the user data, the service product data and the transaction data obtained in each transaction to generate a new vector x ═ c1,c2,…,cx,p1,p2,…,py,t1,t2,…,tz]. In the splicing process, the information of the digital type is directly converted into a vector; uniformly converting the information of the time types into days; for the text type information, the text is converted into word bags by word segmentation, and then the text vectorization is carried out. In this embodiment, a word2vec text vectorization method is preferably adopted.
Dividing the spliced new vector into N fields, i.e. x ═ field1,field2，…,fieldN]Setting definition targets of products corresponding to the new vectors in all transactions, wherein if the products corresponding to the new vectors are purchased, the definition targets are y equal to 1, taking the products which are not purchased in the transactions as a negative example, and the definition targets y equal to 0 of the products corresponding to the new vectors. And divides the new vectors in all transactions into a training set and a validation set.
S2, a neural network and a learnable matrix decomposition model are built, and the data obtained in the step S1 are input to train the model, so that a recommendation model with the best effect is obtained.
The specific method for training the model comprises the following steps:
s2.1 defining an objective function
The method is used for predicting the probability of one transaction generated at a certain moment and determining the size K of a decomposition matrix;
s2.2 two fully-connected networks, i.e. linear layers, are built for each domain of the new vector, where the first fully-connected layer outputs a vector o e R of length 11The second full-connection layer outputs a vector m ∈ R with the length of KKWhen there are N domains, the final output is: o1,o2,…,oNAnd m1,m2,…,mN. Calculating A from the output values of two fully connected networks1，A1The calculation formula of (2) is as follows:
where N is the number of fields, oiIs the output value of the first fully connected network; m isiAnd mjAre output values of the second fully connected network.
A vector o e R with output of length 11And the other outputs a vector m ∈ R with the length of KK。；
S2.3 output value m of the second of the two fully-connected networks1,m2,…,mNThe output vectors M of the second fully-connected network are spliced together and input into a neural network model containing multiple layers of neurons, and the output result of the model is A2(ii) a The neural network model in this embodiment includes a linear layer, an activation layer, Dropout, and Batch normalization.
S2.4 prediction results
According to the prediction result
And defining loss (loss) back propagation generated by the target y, and updating the weight of each full-connection network untilTo obtain the best possible recommendation model.
And after the step S2.4 is finished, inputting the data in the verification set into the recommended model with the best effect, if the verification result is improved, saving the model, otherwise, repeating the steps S2.1-S2.4, and retraining the model.
S3, inputting the actual demand into the recommendation model with the best effect, and obtaining the service transaction recommendation.
S3.1, acquiring an actual demand vector and a recommended scene vector;
s3.2 input service product vector p with number ii1,pi2,…,piy]Generating a vector ziIf the total number of products is n, generating all product vector sets Z ═ Z1,z2,...，zn}；
S3.3, inputting all product vector sets X into the recommendation model with the best effect, and outputting a prediction result set
And S3.4, sorting the prediction results in the prediction result set from big to small, and recommending the results ranked in the front.
Example two
Based on the same inventive concept, the embodiment discloses an intelligent service transaction recommendation system, which comprises:
the data acquisition module is used for acquiring the preprocessed data;
the model training module is used for constructing a neural network and a learnable matrix decomposition model, inputting the data obtained in the data obtaining module to train the model and obtaining a recommendation model with the best effect;
and the interaction module is used for inputting the actual requirements into the recommendation model with the best effect to obtain service transaction recommendation.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims. The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application should be defined by the claims.
Claims (10)
1. An intelligent service transaction recommendation method is characterized by comprising the following steps:
s1, acquiring the preprocessed data;
s2, constructing a neural network and a learnable matrix decomposition model, inputting the data obtained in the step S1 to train the model, and obtaining a recommendation model with the best effect;
s3, inputting the actual demand into the recommendation model with the best effect, and obtaining the service transaction recommendation.
2. The intelligent service transaction recommendation method according to claim 1, wherein the data in step S1 includes user data, service product data and transaction data, and the user data, service product data and transaction data obtained in each transaction are merged to generate a new vector.
3. The intelligent service transaction recommendation method of claim 2, wherein the method for generating the new vector by splicing comprises: for information of digital type, directly converting into vector; uniformly converting the information of the time types into days; for the text type information, the text is converted into word bags by word segmentation, and then the text vectorization is carried out.
4. The intelligent service transaction recommendation method according to claim 2, wherein the preprocessing method of the data in the step S1 is as follows: and dividing the spliced new vectors into N domains, setting definition targets of products corresponding to the new vectors in all transactions, and dividing the new vectors in all transactions into a training set and a verification set.
5. The intelligent service transaction recommendation method of claim 4, wherein the new vector corresponds to a product whose definition target y is 1 if the product has been purchased, otherwise, the new vector corresponds to a product whose definition target y is 0.
6. The intelligent service transaction recommendation method of claim 4, wherein the method for training the model in step S2 is as follows:
s2.1 defining an objective function
The method is used for predicting the probability of one transaction generated at a certain moment and determining the size K of a decomposition matrix;
s2.2, two fully-connected networks are constructed for each domain of the new vector, and A is calculated according to output values of the two fully-connected networks1；
S2.3, inputting an output vector M of the second of the two fully-connected networks into the model, wherein the output result of the model is A2；
S2.4 prediction results
According to the prediction result
And defining an object y, and updating the weight of the fully-connected network until a best-effect recommendation model is obtained.
7. The intelligent service transaction recommendation method according to claim 6, wherein after the step S2.4 is finished, the data in the verification set is input into the recommendation model with the best effect, if the verification result is improved, the model is saved, otherwise, the steps S2.1-S2.4 are repeated, and the model is retrained.
8. The intelligent service transaction recommendation method of claim 6, wherein a in step S2.21The calculation formula of (2) is as follows:
where N is the number of fields, oiIs the output value of the first fully connected network; m isiAnd mjAre output values of the second fully connected network.
9. The intelligent service transaction recommendation method according to claim 7, wherein the step S3 specifically comprises the steps of:
s3.1, acquiring an actual demand vector and a transaction data vector;
s3.2 input service product vector | p with number ii1，pi2，...，piy]Generating a vector xiIf the total number of products is n, generating all product vector sets Z ═ Z1，z2，...，zn}；
S3.3, inputting the vector sets Z of all the products into the recommendation model with the best effect, and outputting a prediction result set
And S3.4, sorting the prediction results in the prediction result set from big to small, and recommending the results ranked in the front.
10. A smart services transaction recommendation system, comprising:
the data acquisition module is used for acquiring the preprocessed data;
the model training module is used for constructing a neural network and a learnable matrix decomposition model, inputting the data obtained in the data obtaining module to train the model and obtaining a recommendation model with the best effect;
and the interaction module is used for inputting the actual requirements into the recommendation model with the best effect to obtain service transaction recommendation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110571801.3A CN113222707B (en) | 2021-05-25 | 2021-05-25 | Intelligent service transaction recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110571801.3A CN113222707B (en) | 2021-05-25 | 2021-05-25 | Intelligent service transaction recommendation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113222707A true CN113222707A (en) | 2021-08-06 |
| CN113222707B CN113222707B (en) | 2024-02-27 |
Family
ID=77098293
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110571801.3A Active CN113222707B (en) | 2021-05-25 | 2021-05-25 | Intelligent service transaction recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113222707B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012150561A (en) * | 2011-01-17 | 2012-08-09 | Nippon Telegr & Teleph Corp <Ntt> | Information recommendation device, method, and program |
| US20180107787A1 (en) * | 2016-10-17 | 2018-04-19 | International Business Machines Corporation | Workflow for automatic measurement of doppler pipeline |
| CN109104416A (en) * | 2018-07-24 | 2018-12-28 | 成都电科致远网络科技有限公司 | A kind of more equipment cooperation cloud service platform based on PON local area network |
| CN110516165A (en) * | 2019-08-28 | 2019-11-29 | 安徽农业大学 | A kind of cross-cutting recommended method of hybrid neural networks based on text UGC |
| US20200059451A1 (en) * | 2018-08-14 | 2020-02-20 | Didi Research America, Llc | System and method for detecting generated domain |
| CN111640483A (en) * | 2020-06-09 | 2020-09-08 | 南京邮电大学 | Fitness scheme recommendation method based on AKC model |
| CN112131471A (en) * | 2020-09-23 | 2020-12-25 | 平安国际智慧城市科技股份有限公司 | Method, device, equipment and medium for relation recommendation based on unauthorized undirected graph |
| CN112182131A (en) * | 2020-09-28 | 2021-01-05 | 中国电子科技集团公司第五十四研究所 | Remote sensing image recommendation method based on multi-attribute fusion |
| CN112395643A (en) * | 2020-11-23 | 2021-02-23 | 中国人民大学 | Data privacy protection method and system of neural network |
| CN112464097A (en) * | 2020-12-07 | 2021-03-09 | 广东工业大学 | Multi-auxiliary-domain information fusion cross-domain recommendation method and system |
| WO2021071711A1 (en) * | 2019-10-09 | 2021-04-15 | The Uab Research Foundation | Method for uncertainty estimation in deep neural networks |
| CN112785376A (en) * | 2021-01-20 | 2021-05-11 | 电子科技大学 | Multi-domain recommendation method based on multi-task learning |
| CN112801125A (en) * | 2019-11-13 | 2021-05-14 | 辉达公司 | Synthesizing data for training one or more neural networks |
| CN112819222A (en) * | 2021-01-29 | 2021-05-18 | 清华大学 | Prediction method and device of taxi taking demand and traffic flow combined prediction model |
| US20210149671A1 (en) * | 2019-11-19 | 2021-05-20 | Intuit Inc. | Data structures and methods for enabling cross domain recommendations by a machine learning model |
-
2021
- 2021-05-25 CN CN202110571801.3A patent/CN113222707B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2012150561A (en) * | 2011-01-17 | 2012-08-09 | Nippon Telegr & Teleph Corp <Ntt> | Information recommendation device, method, and program |
| US20180107787A1 (en) * | 2016-10-17 | 2018-04-19 | International Business Machines Corporation | Workflow for automatic measurement of doppler pipeline |
| CN109104416A (en) * | 2018-07-24 | 2018-12-28 | 成都电科致远网络科技有限公司 | A kind of more equipment cooperation cloud service platform based on PON local area network |
| US20200059451A1 (en) * | 2018-08-14 | 2020-02-20 | Didi Research America, Llc | System and method for detecting generated domain |
| CN110516165A (en) * | 2019-08-28 | 2019-11-29 | 安徽农业大学 | A kind of cross-cutting recommended method of hybrid neural networks based on text UGC |
| WO2021071711A1 (en) * | 2019-10-09 | 2021-04-15 | The Uab Research Foundation | Method for uncertainty estimation in deep neural networks |
| CN112801125A (en) * | 2019-11-13 | 2021-05-14 | 辉达公司 | Synthesizing data for training one or more neural networks |
| US20210149671A1 (en) * | 2019-11-19 | 2021-05-20 | Intuit Inc. | Data structures and methods for enabling cross domain recommendations by a machine learning model |
| CN111640483A (en) * | 2020-06-09 | 2020-09-08 | 南京邮电大学 | Fitness scheme recommendation method based on AKC model |
| CN112131471A (en) * | 2020-09-23 | 2020-12-25 | 平安国际智慧城市科技股份有限公司 | Method, device, equipment and medium for relation recommendation based on unauthorized undirected graph |
| CN112182131A (en) * | 2020-09-28 | 2021-01-05 | 中国电子科技集团公司第五十四研究所 | Remote sensing image recommendation method based on multi-attribute fusion |
| CN112395643A (en) * | 2020-11-23 | 2021-02-23 | 中国人民大学 | Data privacy protection method and system of neural network |
| CN112464097A (en) * | 2020-12-07 | 2021-03-09 | 广东工业大学 | Multi-auxiliary-domain information fusion cross-domain recommendation method and system |
| CN112785376A (en) * | 2021-01-20 | 2021-05-11 | 电子科技大学 | Multi-domain recommendation method based on multi-task learning |
| CN112819222A (en) * | 2021-01-29 | 2021-05-18 | 清华大学 | Prediction method and device of taxi taking demand and traffic flow combined prediction model |
Non-Patent Citations (6)
| Title |
|---|
| ANU, HA: "An Approach to Recommendation of Verbosity Log Levels Based on Logging Intention", 《 2019 IEEE INTERNATIONAL CONFERENCE ON SOFTWARE MAINTENANCE AND EVOLUTION 》, 10 April 2020 (2020-04-10), pages 125 - 134 * |
| XIANGNAN HE: "Neural Collaborative Filtering", 《PROCEEDINGS OF THE 26TH INTERNATIONAL CONFERENCEON WORLD WIDE WEB COMPANION. PERTH:INTERNATIONAL WORLD WIDE WEB CONFERENCES STEERING COMMITTE》, pages 173 - 182 * |
| ZHANG, WEI: "Pilot-Assisted MIMO-V-OFDM Systems: Compressed Sensing and Deep Learning Approaches", 《IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC》, vol. 8, 28 April 2020 (2020-04-28), pages 7142 - 7159, XP011765573, DOI: 10.1109/ACCESS.2020.2964046 * |
| 曲彦儒: "面向多域离散数据的神经网络框架", 《中国优秀硕士学位论文全 文数据库 信息科技辑》, 30 June 2020 (2020-06-30), pages 140 - 101 * |
| 秦波: "比特币与法定数字货币", 《密码学报》, no. 4, 30 April 2017 (2017-04-30) * |
| 秦若熙: "基于深度网络的多模态医学数据表征学习技术研究", 《中国优秀硕士学位论文全 文数据库 信息科技辑》, 28 February 2021 (2021-02-28), pages 138 - 955 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113222707B (en) | 2024-02-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200050941A1 (en) | Machine learning systems and methods for attributed sequences | |
| Lee et al. | Deep learning based recommender system using cross convolutional filters | |
| US20090112752A1 (en) | Distributed Scoring of Data Transactions | |
| Liu et al. | Interval-valued intuitionistic fuzzy ordered weighted cosine similarity measure and its application in investment decision-making | |
| CN110619540A (en) | Click stream estimation method of neural network | |
| CN110321473B (en) | Multi-modal attention-based diversity preference information pushing method, system, medium and device | |
| CN110796499A (en) | Advertisement conversion rate estimation model and training method thereof | |
| Lu et al. | Learning from multi-view multi-way data via structural factorization machines | |
| CN110727855A (en) | Personalized recommendation method based on improved factorization machine | |
| CN111382843A (en) | Method and device for establishing upstream and downstream relation recognition model of enterprise and relation mining | |
| CN116629977A (en) | Electronic commerce decision method and system based on big data | |
| Zhang et al. | Prediction of repeat customers on e-commerce platform based on blockchain | |
| CN113298546A (en) | Sales prediction method and device, and commodity processing method and device | |
| Liu | E-commerce precision marketing model based on convolutional neural network | |
| CN110826686A (en) | Machine learning system and method with attribute sequence | |
| Vizuete-Luciano et al. | Decision making in the assignment process by using the Hungarian algorithm with OWA operators | |
| CN110213239B (en) | Suspicious transaction message generation method and device and server | |
| CN112348590A (en) | Method and device for determining value of article, electronic equipment and storage medium | |
| CN116910621A (en) | Decision support system and method for textile chemical fiber supply chain management | |
| CN113222707B (en) | Intelligent service transaction recommendation method and system | |
| CN116127049A (en) | Model training method, text generation method, terminal device and computer medium | |
| CN111046280B (en) | Cross-domain recommendation method applying FM | |
| Wu et al. | Symphony in the latent space: provably integrating high-dimensional techniques with non-linear machine learning models | |
| CN111626331B (en) | Automatic industry classification device and working method thereof | |
| CN114092188A (en) | Recommendation system algorithm of lightweight B2B E-commerce platform |
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 |