CN115019933A - Amblyopia training scheme recommendation method fusing GMF and CDAE - Google Patents
Amblyopia training scheme recommendation method fusing GMF and CDAE Download PDFInfo
- Publication number
- CN115019933A CN115019933A CN202210682588.8A CN202210682588A CN115019933A CN 115019933 A CN115019933 A CN 115019933A CN 202210682588 A CN202210682588 A CN 202210682588A CN 115019933 A CN115019933 A CN 115019933A
- Authority
- CN
- China
- Prior art keywords
- model
- user
- layer
- cdae
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000012549 training Methods 0.000 title claims abstract description 85
- 201000009487 Amblyopia Diseases 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 29
- 230000004927 fusion Effects 0.000 claims abstract description 16
- 230000006870 function Effects 0.000 claims abstract description 13
- 239000013598 vector Substances 0.000 claims description 36
- 239000011159 matrix material Substances 0.000 claims description 20
- 230000009467 reduction Effects 0.000 claims description 19
- 238000000354 decomposition reaction Methods 0.000 claims description 18
- 230000003993 interaction Effects 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 11
- 238000013507 mapping Methods 0.000 claims description 7
- 230000004913 activation Effects 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 230000014509 gene expression Effects 0.000 claims description 3
- 230000003213 activating effect Effects 0.000 claims description 2
- 238000012937 correction Methods 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 description 5
- 230000004438 eyesight Effects 0.000 description 4
- 230000001149 cognitive effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 230000004382 visual function Effects 0.000 description 2
- 208000004350 Strabismus Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 208000030533 eye disease Diseases 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000036403 neuro physiology Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000638 stimulation Effects 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 238000002645 vision therapy Methods 0.000 description 1
- 230000009978 visual deterioration Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/30—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to physical therapies or activities, e.g. physiotherapy, acupressure or exercising
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Physical Education & Sports Medicine (AREA)
- Epidemiology (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A recommendation method for amblyopia training scheme fusing GMF and CDAE belongs to the technical field of computers. It comprises the following steps: 1. acquiring user information data in a weak training platform database; 2. respectively extracting data characteristics of input data through a CDAE model and a GMF model; 3. the fusion layer performs weight fusion on the outputs of the CDAE model and the GMF model; 4. and optimizing the function learning model parameters to minimize the prediction error of the prediction score and the real score, and outputting a user recommendation result. The invention introduces intelligent recommendation into the amblyopia training system, solves the artificial formula by utilizing the neural network algorithm in the field of computers, and liberates medical personnel and technical personnel; the CDAE model and the GMF model are introduced, so that the problem of insufficient recommendation accuracy caused by data sparsity is effectively solved, and the recommendation efficiency is improved.
Description
Technical Field
The invention belongs to the technical field of computers, relates to a neural collaborative filtering and recommending system, and particularly relates to a multimedia amblyopia training scheme recommending method integrating generalized matrix decomposition and a simultaneous noise reduction automatic encoder.
Background
Amblyopia is a developmental ophthalmic disease in children, due to the abnormal visual experience during the developmental stage of vision, with the best corrected visual deterioration of one or both eyes, without organic eye diseases.
With the development of modern internet information technology, amblyopia treatment methods are diversified. The multimedia amblyopia training system is widely developed. The multimedia amblyopia training system comprehensively uses the theory and method of neurobiology, psychophysics and computer vision, utilizes the plasticity of the brain nervous system, combines amblyopia treatment with computer games, and improves the visual function through various biological stimulation. The multimedia amblyopia training system adds more interestingness, and the training mode combines the cognitive training content to improve the training effect.
The existing amblyopia training system achieves the purpose of amblyopia treatment by providing a series of training items for users, and the training items mostly use elements such as animation, graphics and the like as carriers to stimulate human eyes. Users in such multimedia amblyopia systems need to complete one or more session cycles of training, with each training session requiring multiple training sessions. In recent years, amblyopia training software has been studied more and more, and as early as 2003, Cecilia et al developed three multimedia programs for treating children's vision. In 2005, king et al combined computer technology with modern neurophysiology and psychology of children to achieve a multimedia amblyopia treatment system for children. Lukasz et al developed a computer therapy system for strabismus. Li Xiangjie designs and realizes a self-adaptive training system for children, and parameter self-adaptive strategies including difficulty and grade are provided by considering parameters in each training item. In yew, an online amblyopia training system based on cognitive theory is designed, and the patient compliance is improved by combining the cognitive theory. Therefore, the multimedia amblyopia training system is widely applied at present, and the research and the improvement on the amblyopia training system have great significance for the rehabilitation of the amblyopia patient.
In these multimedia amblyopia training systems, a doctor or a professional specifies a training scheme for a patient, and these training items are not changed after being specified until the user completes course training. The training scheme designated manually has the defects of singleness, limitation and the like, so that the training scheme cannot be adjusted in time in the training process of the user, and the user loses training interest due to repeated training and cannot achieve the optimal training effect. At present, the systems only manage basic information of patients, do not collect data in the training process of the patients, and cannot track the training condition of the patients in time. Therefore, the formula operation not only needs to spend more manpower, but also can not achieve better training effect because the appointed training items can not be adjusted in time in the training process of the patient.
The recommendation based on the neural collaborative filtering is widely applied in recent years, and in order to improve the effectiveness of a user training project, reduce unnecessary time loss and relieve the problem that manual formulation of doctors and professionals consumes manpower, the recommendation of a neural network is introduced, so that effective training scheme recommendation is realized.
Disclosure of Invention
The invention provides a recommendation method of an amblyopia training scheme, which integrates generalized matrix decomposition and a collaborative noise reduction automatic encoder, aiming at the problems of limitation and singleness of a manual appointed scheme, poor user training effect and the like in an amblyopia training system.
The invention provides the following technical scheme: a recommendation system of an amblyopia training scheme fusing GMF and CDAE comprises an input layer, a hidden layer, a fusion layer and an output layer; the hidden layer comprises a generalized matrix decomposition GMF model and a collaborative noise reduction automatic encoder CDAE model, the collaborative noise reduction automatic encoder CDAE adopts a three-layer network and two full-connection layers, and an offset vector and a tanh activating function are used;
the recommendation method based on the recommendation system specifically comprises the following steps:
step 2, respectively extracting data characteristics of input data by a collaborative noise reduction automatic encoder CDAE model and a generalized matrix decomposition GMF model in the hidden layer;
step 3, the fusion layer performs weight fusion on the outputs of the CDAE model and the GMF model of the collaborative noise reduction automatic encoder in the hidden layer so as to balance high and low-order characteristics; obtaining predicted user-training item scores through fusion layers
Transmitting it to an output layer for feedback optimization;
step 4, the output layer adopts mean square error loss MSE as a target optimization function learning model parameter to minimize the prediction error of the prediction score and the real score, and the model parameter is subjected to iterative correction through an AdamW optimizer; and finally, selecting the set TOP-N recommendation number to recommend the user according to the high-low sequence of the prediction scores.
Further, in the step 1, mapping data input by an input layer into a hidden vector describing the user and the item in the context of the latent factor model, and then taking the hidden vector as the input of the generalized matrix decomposition GMF model; combining the user hidden vector which is the same as the input of the generalized matrix decomposition GMF model with damaged user-project interaction data to serve as the input of a collaborative noise reduction automatic encoder CDAE model, and randomly setting 0 to the data which is 1 in the user-training project scoring vector to generate the damaged data in the data processing process.
Further, in the CDAE model, the user-training item data from the input layer is input to the two fully-connected layers to extract the high-order interaction information of the user-training item; an activation function is introduced into the two fully-connected layers to realize a nonlinear process, which specifically comprises the following steps:
firstly, mapping the processed damage vector in the input layer to a first layer of a full connection layer; the feature expressions of the first-level fully-connected layer are then mapped to the second-level fully-connected layer to reconstruct the input user-item score vector.
Further, it is characterized byIn the generalized matrix decomposition GMF model, firstly, corresponding elements of a user hidden vector and a training item hidden vector are multiplied, and then a vector obtained after multiplication is multiplied by a weight W G Then adding, and inputting the result into the activation function for further mapping to introduce a nonlinear process.
By adopting the technology, compared with the prior art, the invention has the following beneficial effects:
1) the invention introduces intelligent recommendation into the amblyopia training system, solves the artificial formula by utilizing the neural network algorithm in the field of computers, and liberates medical personnel and technical personnel;
2) in the invention, a collaborative noise reduction automatic encoder is introduced to carry out damage processing on data and extract high-order characteristics of the data, so that the problem of insufficient recommendation accuracy rate caused by data sparsity can be effectively solved;
3) according to the invention, the collaborative noise reduction automatic encoder and the generalized matrix decomposition model are combined, the high-order nonlinear characteristic and the low-order linear characteristic of the data are extracted, and the recommendation efficiency is improved.
Drawings
FIG. 1 is a schematic structural view of the present invention;
FIG. 2 is a schematic structural view of the present invention;
FIG. 3 is a schematic view of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
On the contrary, the invention is intended to cover alternatives, modifications, equivalents and alternatives which may be included within the spirit and scope of the invention as defined by the appended claims. Furthermore, in the following detailed description of the present invention, certain specific details are set forth in order to provide a better understanding of the present invention. It will be apparent to one skilled in the art that the present invention may be practiced without these specific details.
Referring to fig. 1-3, a recommendation method for amblyopia training scheme fusing GMF and CDAE is based on a recommendation system for amblyopia training scheme, which includes an input layer for data processing, a hidden layer composed of a CDAE model and a GMF model, a fusion layer for result fusion, and an output layer.
Specifically, the input layer:
(1) and acquiring user personal basic information and various visual function inspection information in the amblyopia training platform database, amblyopia training item related information and historical training scoring records from the rear end. The user personal basic information comprises a user ID, a name, a sex, an age, corrected eyesight, a symptom label and the like, the amblyopia training item information comprises a training item ID, an item name, a strategy ID, a strategy name, a category ID and the like, the historical training scoring record comprises the user ID, the training item ID and the user scoring.
(2) The GMF part maps the data of the input layer into hidden vectors describing the user and the item in the context of the latent factor model, and then takes this as input.
(3) The CDAE component takes as input a user hidden vector similar to that in the GMF input in combination with broken user-item interaction data. When the input of the CDAE part is processed, the data of 1 in the user-item score vector is randomly set to 0 to generate broken data, as shown in fig. 2.
The specific formula is as follows:
where p is broken user-item interaction data
Middle y ui Probability of not being 0, y ui In order for the user to score the training program,
for after treatmentQ is a random probability, and σ is 1/(1-q).
Specifically, the hidden layer:
(1) the CDAE inputs the user-training item data from the input layer to the two fully connected layers to extract the high-order interaction information of the user-training item. And introducing an activation function into the two fully-connected layers to realize a nonlinear process.
The method comprises the following specific steps:
first, the processed damage vector in the input layer is mapped to the first layer of the fully-connected layer. The formula is as follows:
W C representing a weight matrix, V, between an input level node and a CDAE first level node u A hidden vector representing a user input node. V u Is a user-specific vector, i.e. for each user there is a unique vector V u . b is the offset vector and h (-) is an element-based mapping function.
The feature expressions of the first-level fully-connected layer are then mapped to the second-level fully-connected layer to reconstruct an input user-item score vector, and a user-item score is predicted for each of the reconstructed vectors, as calculated by:
(2) the GMF model does not directly use the linear inner product of the user implicit vector and the training project implicit vector as a prediction score, firstly, the user implicit vector and corresponding elements of the training project implicit vector are multiplied, and then the vector obtained after multiplication is multiplied by the weight W G Then adding, and inputting the result into the activation function for further mapping to introduce a nonlinear process.
The specific formula is as follows:
P u and Q i Implicit vectors representing user U and item I, respectively
Specifically, the fusion layer: the output result of CDAE and the output result of GMF in the hidden layer are weighted by corresponding weights
The fusion formula is:
complete user-project prediction scores through fusion layer
The incoming output layer is used for feedback optimization.
Specifically, the output layer: the output layer adopts mean square error loss (MSE) as a target optimization function learning model parameter to minimize the prediction error of the prediction score and the real score.
The specific formula is as follows:
n represents the number of items. The model parameters were iteratively corrected by an AdamW optimizer. And finally, selecting the set TOP-N recommendation number according to the high-low order of the prediction scores to recommend the user, wherein a detailed flow chart of specific recommendation is shown in FIG. 3.
The invention provides a multimedia amblyopia training scheme recommendation method fusing generalized matrix decomposition and a collaborative noise reduction automatic encoder. The method effectively utilizes a collaborative noise reduction automatic encoder CDAE and a generalized matrix decomposition GMF model, based on a large amount of user training data obtained by doctor-guided training, user information is combined with user-training item interaction data randomly set to zero at a certain probability to serve as input, and implicit feedback data is adopted to model the preference of a user, so that the problem that the recommendation accuracy is insufficient due to the fact that the system is difficult to accurately extract the characteristics of the data due to the sparsity of the user-training item interaction data can be effectively solved.
The method comprises the steps of firstly learning high-order and low-order interactive features implicit between a user and a training project through a collaborative noise reduction automatic encoder and generalized matrix decomposition, and then performing weighted fusion before an output layer. The deep interaction relation between the user and the training items is more comprehensively learned, and the recommendation performance is further improved. The invention randomly zeros the data by the corresponding probability value of the user and the training item matrix to carry out the breakage, inputs the processed breakage data into the CDAE model to train, and inputs the user and the training item without breakage into the GMF model to train. And finally, weighting and fusing the output results of the two models by a certain weight to balance the high-order and low-order characteristics. And obtaining a prediction score by combining the preference information, and recommending the user vision training scheme according to the prediction score. Therefore, the problem that a large amount of human resources are consumed due to manual configuration schemes is simplified, intelligent recommendation and user-friendly use are achieved, and training effects are improved.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (4)
1. A recommendation method for amblyopia training scheme fusing GMF and CDAE is characterized by comprising the following steps: the recommendation system comprises an input layer, a hidden layer, a fusion layer and an output layer; the hidden layer comprises a generalized matrix decomposition GMF model and a collaborative noise reduction automatic encoder CDAE model, the collaborative noise reduction automatic encoder CDAE adopts a three-layer network and two full-connection layers, and an offset vector and a tanh activating function are used;
the recommendation method based on the recommendation system specifically comprises the following steps:
step 1, an input layer acquires user information data in a disadvantaged training platform database, wherein the user information data comprises user personal basic information, amblyopia training project related information and user-training project interaction information; selecting a part of data to carry out damage processing, inputting the processed damaged data into a CDAE (collaborative noise reduction) model of an automatic encoder, and inputting the other part of undamaged data into a GMF (generalized matrix decomposition) model;
step 2, respectively extracting data characteristics of input data by a collaborative noise reduction automatic encoder CDAE model and a generalized matrix decomposition GMF model in the hidden layer;
step 3, the fusion layer performs weight fusion on the outputs of the CDAE model and the GMF model of the collaborative noise reduction automatic encoder in the hidden layer so as to balance high and low-order characteristics; obtaining predicted user-training item scores through fusion layers
Transmitting it to an output layer for feedback optimization;
step 4, the output layer adopts mean square error loss MSE as a target optimization function learning model parameter to minimize the prediction error of the prediction score and the real score, and the model parameter is subjected to iterative correction through an AdamW optimizer; and finally, selecting the set TOP-N recommendation number to recommend the user according to the high-low sequence of the prediction scores.
2. The method according to claim 1, wherein in step 1, the data input by the input layer is mapped into hidden vectors describing the user and the item in the context of the latent factor model, and then the hidden vectors are used as the input of the generalized matrix decomposition GMF model; combining the user hidden vector which is the same as the input of the generalized matrix decomposition GMF model with damaged user-project interaction data to serve as the input of a collaborative noise reduction automatic encoder CDAE model, and randomly setting 0 to the data which is 1 in the user-training project scoring vector to generate the damaged data in the data processing process.
3. The method according to claim 2, wherein in the CDAE model, the user-training item data from the input layer is input to two fully-connected layers to extract the high-order interaction information of the user-training item; an activation function is introduced into the two fully-connected layers to realize a nonlinear process, which specifically comprises the following steps:
firstly, mapping the processed damage vector in the input layer to a first layer of a full connection layer; the feature expressions of the first-level fully-connected layer are then mapped to the second-level fully-connected layer to reconstruct the input user-item score vector.
4. The method as claimed in claim 3, wherein in the GMF model with generalized matrix decomposition, the hidden vector of the user is multiplied by the corresponding element of the hidden vector of the training item, and the multiplied vector is multiplied by the weight W G Then adding, and inputting the result into the activation function for further mapping to introduce a nonlinear process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210682588.8A CN115019933B (en) | 2022-06-16 | 2022-06-16 | Weak vision training scheme recommendation method integrating GMF and CDAE |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210682588.8A CN115019933B (en) | 2022-06-16 | 2022-06-16 | Weak vision training scheme recommendation method integrating GMF and CDAE |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115019933A true CN115019933A (en) | 2022-09-06 |
| CN115019933B CN115019933B (en) | 2024-06-21 |
Family
ID=83074365
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210682588.8A Active CN115019933B (en) | 2022-06-16 | 2022-06-16 | Weak vision training scheme recommendation method integrating GMF and CDAE |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115019933B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116775996A (en) * | 2023-06-21 | 2023-09-19 | 广州视景医疗软件有限公司 | Visual training project recommending method and device based on user feedback |
| CN117809807A (en) * | 2024-01-22 | 2024-04-02 | 中科网联(武汉)信息技术有限公司 | Visual training method, system and storage medium based on interaction platform |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109783739A (en) * | 2019-01-23 | 2019-05-21 | 北京工业大学 | A kind of collaborative filtering recommending method based on the sparse noise reduction self-encoding encoder enhancing of stacking |
| CN110059262A (en) * | 2019-04-19 | 2019-07-26 | 武汉大学 | A kind of construction method and device, item recommendation method of the project recommendation model based on hybrid neural networks |
| CN110807154A (en) * | 2019-11-08 | 2020-02-18 | 内蒙古工业大学 | Recommendation method and system based on hybrid deep learning model |
| CN110851705A (en) * | 2019-10-09 | 2020-02-28 | 天津大学 | Project-based collaborative storage recommendation method and recommendation device thereof |
| WO2020233245A1 (en) * | 2019-05-20 | 2020-11-26 | 山东科技大学 | Method for bias tensor factorization with context feature auto-encoding based on regression tree |
-
2022
- 2022-06-16 CN CN202210682588.8A patent/CN115019933B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109783739A (en) * | 2019-01-23 | 2019-05-21 | 北京工业大学 | A kind of collaborative filtering recommending method based on the sparse noise reduction self-encoding encoder enhancing of stacking |
| CN110059262A (en) * | 2019-04-19 | 2019-07-26 | 武汉大学 | A kind of construction method and device, item recommendation method of the project recommendation model based on hybrid neural networks |
| WO2020233245A1 (en) * | 2019-05-20 | 2020-11-26 | 山东科技大学 | Method for bias tensor factorization with context feature auto-encoding based on regression tree |
| CN110851705A (en) * | 2019-10-09 | 2020-02-28 | 天津大学 | Project-based collaborative storage recommendation method and recommendation device thereof |
| CN110807154A (en) * | 2019-11-08 | 2020-02-18 | 内蒙古工业大学 | Recommendation method and system based on hybrid deep learning model |
Non-Patent Citations (2)
| Title |
|---|
| 刘振鹏;尹文召;王文胜;孙静薇;: "HRS-DC：基于深度学习的混合推荐模型", 计算机工程与应用, no. 14, 13 June 2019 (2019-06-13) * |
| 张杰;付立军;刘俊明;: "基于混合自编码器的协同过滤推荐算法优化", 计算机系统应用, no. 05, 15 May 2019 (2019-05-15) * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116775996A (en) * | 2023-06-21 | 2023-09-19 | 广州视景医疗软件有限公司 | Visual training project recommending method and device based on user feedback |
| CN117809807A (en) * | 2024-01-22 | 2024-04-02 | 中科网联(武汉)信息技术有限公司 | Visual training method, system and storage medium based on interaction platform |
| CN117809807B (en) * | 2024-01-22 | 2024-05-31 | 中科网联(武汉)信息技术有限公司 | Visual training method, system and storage medium based on interaction platform |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115019933B (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115019933A (en) | Amblyopia training scheme recommendation method fusing GMF and CDAE | |
| Arbib et al. | Synthetic brain imaging: grasping, mirror neurons and imitation | |
| Tam et al. | A minimal set of electrodes for motor imagery BCI to control an assistive device in chronic stroke subjects: a multi-session study | |
| Hussien et al. | EEG channel selection using a modified grey wolf optimizer | |
| Di Lascio et al. | A fuzzy-based methodology for the analysis of diabetic neuropathy | |
| CN103226665B (en) | Based on the human brain health exercising system of SAAS platform | |
| Vakoch et al. | The evolution of psychotherapy training: Reflections on manual‐based learning and future alternatives | |
| Kabari et al. | Neural networks and decision trees for eye diseases diagnosis | |
| Shaffer et al. | Manufacturing multiple meanings of addiction: Time-limited realities | |
| Galdi et al. | Gross-motor skills for potential intelligence descriptive study in a kindergarten | |
| Kadhim | FNDSB: A fuzzy-neuro decision support system for back pain diagnosis | |
| Han et al. | Noise-Factorized Disentangled Representation Learning for Generalizable Motor Imagery EEG Classification | |
| Stokes et al. | Naturalistic approaches to creativity | |
| Zhang et al. | [Retracted] Swimming Training Evaluation Method Based on Convolutional Neural Network | |
| CN114052734B (en) | Electroencephalogram emotion recognition method based on progressive graph convolution neural network | |
| Norton | Motivational interviewing: an appropriate model for engaging clients with drug abuse problems in the probation service | |
| Oliva et al. | Cognitive Society. | |
| CN118471488B (en) | Diagnosis and treatment auxiliary device and diagnosis and treatment auxiliary method based on large language model | |
| Zhong | Motor Imagery-based BCI Perspective in Procedure and Analysis | |
| US20230360764A1 (en) | Personalized exercise guidance system and method based on machine learning | |
| Li et al. | The SSHVEP Paradigm-Based Brain Controlled Method for Grasping Robot Using MVMD Combined CNN Model | |
| MO | Novel Implementation of Cardio-vascular disease (CVD) using machine learning techniques | |
| Li et al. | A decision support system for upper limb rehabilitation robot based on hybrid reasoning with RBR and CBR | |
| Ma et al. | Construction of Neural Network Model for Exercise Load Monitoring Based on Yoga Training Data and Rehabilitation Therapy | |
| Chen et al. | Analysis and evaluation of the effect of sports on promoting health in Traditional Wushu |
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 |