CN111914082A - Online knowledge aggregation method based on SOM neural network algorithm - Google Patents
Online knowledge aggregation method based on SOM neural network algorithm Download PDFInfo
- Publication number
- CN111914082A CN111914082A CN201910387629.9A CN201910387629A CN111914082A CN 111914082 A CN111914082 A CN 111914082A CN 201910387629 A CN201910387629 A CN 201910387629A CN 111914082 A CN111914082 A CN 111914082A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- neural network
- neurons
- weight
- network algorithm
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 17
- 230000002776 aggregation Effects 0.000 title claims abstract description 12
- 238000004220 aggregation Methods 0.000 title claims abstract description 12
- 210000002569 neuron Anatomy 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 5
- 210000002364 input neuron Anatomy 0.000 claims description 2
- 210000004205 output neuron Anatomy 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 abstract 1
- 238000013467 fragmentation Methods 0.000 description 7
- 238000006062 fragmentation reaction Methods 0.000 description 7
- 230000008520 organization Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 239000012634 fragment Substances 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 230000008521 reorganization Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- 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
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to an online knowledge aggregation model based on an SOM neural network algorithm, which comprises the following steps: step 1, acquiring a knowledge element; step 2, setting a neural network variable; step 3, initializing a weight value; step 4, selecting winning neurons; step 5, updating the weight; step 6, updating parameters of the nodes; and 7, outputting the knowledge cluster. The method utilizes the technologies such as the SOM neural network and the like to aggregate the knowledge elements with the flow relation into one knowledge cluster, and integrates the knowledge cluster into a series of knowledge subjects, so that the knowledge subjects are more in line with the requirements of online learning in content and form.
Description
Technical Field
The invention belongs to the cross field of artificial intelligence and online education, and relates to an online knowledge aggregation method based on a self-organizing map (SOM) neural network algorithm.
Background
The fragmented knowledge enables learners to conveniently acquire the knowledge and has some problems. The fragmentation knowledge source is scattered, the integration of knowledge is lacked, and the existence form of the fragmentation knowledge source is random, dynamic and fragmented; secondly, the fragmentized knowledge is easy to cause disorder of knowledge structure, and the original knowledge structure is damaged in the fragmentizing process of the original knowledge;
therefore, the existing fragmentation knowledge form has the following defects: (1) fragmentation knowledge existence form fragmentation and organization mode disorder. (2) There is a lack of necessary correlation between the fragmentation knowledge.
At present, the organization mode of the fragmentation knowledge is still the traditional and stacked organization mode, and the particles are too coarse. In the fragmented learning process, knowledge recombination based on the prior art cannot be carried out, reorganization methods such as normalized extraction, semantic association and dynamic aggregation are lacked for knowledge fragments, and problems such as knowledge confusion and cognition overload of learners are easily caused;
disclosure of Invention
The invention aims to overcome the defects of the prior art and provides an online knowledge aggregation method based on an SOM neural network algorithm, which is favorable for systematizing a fragmented knowledge structure and improving the acquisition rate of online knowledge.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows: an online knowledge method based on an SOM neural network algorithm comprises the following steps:
step A: acquiring a knowledge element: suppose X is a meta-sample of knowledge that the student is currently learning, P1,P2, P3,…,PnIs an n-dimensional input of a sample of the knowledge elements in the self-organizing neural network, denoted as X ═ P1,P2, P3,…,Pn]
And B: setting neuron variables: z1,Z2,Z3,…,ZmM × n weight coefficient vectors for input neurons and output neurons, denoted as Zj=[P1j,P2j,P3j,…,Pnj]Where j is 1, 2, …, and m is m classes, defining the iteration number num is a.
And C: initialization: the weight of each node is initialized and a value between 0 and 1 is randomly generated. And normalizing the input vector and the weight.
X' ═ X/X; x is the Euclidean norm of the input sample vector
Zj ═ Zj/z; z is the Euclidean norm of the weight vector
Step D: selecting a winning neuron: calculating Euclidean distance D between sample and weight vectorjComparing the sizes between themThe least distant neuron wins the competition and is the winning neuron i (x).
Step E: updating the weight: and updating the neurons in the topological neighborhood of the winning neuron, and normalizing the learned weight again. Order SijRepresenting the distance between nodes i and j, and for i (x) neighbors, assigning them an update weight:
step F: updating parameters of the nodes: updating is performed according to a gradient descending method until convergence.
Δzji=η(t)·Tj,I(x)(t)·(xi-zji)
zji=zji-r*Δzij
η (t): is the inverse of the number of iterations num
Step G: and outputting a result: PT1,PT2,PT3,…,PTnIs the output of a knowledge cluster of n neurons of the neural network, ifThe PTx knowledge cluster is output.
Compared with the prior art, the invention has the advantages and positive effects that:
the method extracts fragmented knowledge into knowledge elements, aggregates the knowledge elements with association relationship into knowledge clusters, and ensures that the relationship between the knowledge elements in the knowledge clusters is more definite and the knowledge structure is more compact, so that the method better meets the requirements of fragmented learning on content and form.
Drawings
FIG. 1 is a SOM neural network topology simulation knowledge cluster aggregation process
Detailed Description
Step A: acquiring a knowledge element: quadruplet representation knowledge element model
The dimension of X is the attribute dimension of the knowledge element sample, and A: ontology structure of knowledge elements, B: concept of a knowledge element, C: one attribute in the concept of a knowledge element, D: a method of concept of a knowledge element.
Taking the knowledge element which is being learned by the user as a sample in a sample set; a1, b1, c1 and d1 represent attribute values of the knowledge elements;
and B: setting neuron variables:z is a weight matrix of 2 x 4, the rows represent the number of classes and the columns represent the dimensions.
Z1,Z2Is a matrix of 1 x 4, Zj=[P1j,P2j,P3j,P4j]J is 1, 2; the iteration number num is 2000;
and C: initialization: the weight of each node is initialized and a value between 0 and 1 is randomly generated.
Z=rand(2,4)
Step D: selecting a winning neuron:
if D is1>D2Then Z for the input sample X2And finally obtaining the winning neurons of all samples as the winning neurons I (x).
Step F: updating parameters of the nodes:
η(t)=1/num
Δzji=η(t)·Tj,I(x)(t)·(xi-zji)
zji=zji-r*Δzij
step G: PT1、PT2Is the output of a knowledge cluster of n neurons of the neural network, ifThen output PT1And (4) knowledge clustering.
The above description is only a preferred embodiment of the present invention, and all changes and modifications that come within the scope of the invention as defined by the appended claims fall within the scope of the invention.
Claims (5)
1. An online knowledge aggregation method based on an SOM neural network algorithm is characterized by comprising the following steps:
step A: acquiring a knowledge element: suppose X is a meta-sample of knowledge that the student is currently learning, P1,P2,P3,…,PnIs an n-dimensional input of a sample of the knowledge elements in the self-organizing neural network, denoted as X ═ P1,P2,P3,…,Pn]
And B: setting neuron variables: z1,Z2,Z3,…,ZmM × n weight coefficient vectors for input neurons and output neurons, denoted as Zj=[P1j,P2j,P3j,…,Pnj]Where j is 1, 2, …, and m is m classes, defining the iteration number num is a.
And C: initialization: the weight of each node is initialized and a value between 0 and 1 is randomly generated. And normalizing the input vector and the weight.
Step D: selecting a winning neuron: calculating Euclidean distance D between sample and weight vectorjAnd comparing the sizes of the neurons, winning the competition of the neurons with the smallest distance, and obtaining the winning neuron I (x).
Step E: updating the weight: and updating the neurons in the topological neighborhood of the winning neuron, and normalizing the learned weight again.
Step F: updating parameters of the nodes: updating is performed according to a gradient descending method until convergence.
2. The SOM neural network algorithm-based online knowledge aggregation method according to claim 1,
in the step A, the knowledge element model is represented by a quadruple,the dimension of X is the attribute dimension of the knowledge element sample, and A: ontology structure of knowledge elements, B: concept of a knowledge element, C: one attribute in the concept of a knowledge element, D: a method of concept of a knowledge element.
3. The SOM neural network algorithm-based online knowledge aggregation method according to claim 1,
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910387629.9A CN111914082A (en) | 2019-05-08 | 2019-05-08 | Online knowledge aggregation method based on SOM neural network algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910387629.9A CN111914082A (en) | 2019-05-08 | 2019-05-08 | Online knowledge aggregation method based on SOM neural network algorithm |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111914082A true CN111914082A (en) | 2020-11-10 |
Family
ID=73242505
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910387629.9A Pending CN111914082A (en) | 2019-05-08 | 2019-05-08 | Online knowledge aggregation method based on SOM neural network algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111914082A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102024179A (en) * | 2010-12-07 | 2011-04-20 | 南京邮电大学 | Genetic algorithm-self-organization map (GA-SOM) clustering method based on semi-supervised learning |
CN102789593A (en) * | 2012-06-18 | 2012-11-21 | 北京大学 | Intrusion detection method based on incremental GHSOM (Growing Hierarchical Self-organizing Maps) neural network |
CN106228274A (en) * | 2016-08-03 | 2016-12-14 | 河海大学常州校区 | Photovoltaic power station power generation amount Forecasting Methodology based on SOM Neural Network Data clustering recognition |
CN107220671A (en) * | 2017-05-27 | 2017-09-29 | 重庆大学 | A kind of Artificial Olfactory on-line correction sample generating method based on self organization map |
-
2019
- 2019-05-08 CN CN201910387629.9A patent/CN111914082A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102024179A (en) * | 2010-12-07 | 2011-04-20 | 南京邮电大学 | Genetic algorithm-self-organization map (GA-SOM) clustering method based on semi-supervised learning |
CN102789593A (en) * | 2012-06-18 | 2012-11-21 | 北京大学 | Intrusion detection method based on incremental GHSOM (Growing Hierarchical Self-organizing Maps) neural network |
CN106228274A (en) * | 2016-08-03 | 2016-12-14 | 河海大学常州校区 | Photovoltaic power station power generation amount Forecasting Methodology based on SOM Neural Network Data clustering recognition |
CN107220671A (en) * | 2017-05-27 | 2017-09-29 | 重庆大学 | A kind of Artificial Olfactory on-line correction sample generating method based on self organization map |
Non-Patent Citations (1)
Title |
---|
何丽: ""基于Web挖掘的决策支持***模型研究"", 《知网》, 15 February 2007 (2007-02-15), pages 3 - 5 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110175386B (en) | Method for predicting temperature of electrical equipment of transformer substation | |
CN106991440B (en) | Image classification method of convolutional neural network based on spatial pyramid | |
CN112087447B (en) | Rare attack-oriented network intrusion detection method | |
CN110084610B (en) | Network transaction fraud detection system based on twin neural network | |
CN110765582B (en) | Self-organization center K-means microgrid scene division method based on Markov chain | |
CN108898213A (en) | A kind of adaptive activation primitive parameter adjusting method towards deep neural network | |
CN111461907A (en) | Dynamic network representation learning method oriented to social network platform | |
CN115758246A (en) | Non-invasive load identification method based on EMD and AlexNet | |
CN114785824B (en) | Intelligent Internet of things big data transmission method and system | |
CN108520986A (en) | A kind of power battery method for group matching based on generation confrontation network | |
CN104850864A (en) | Unsupervised image recognition method based on convolutional neural network | |
CN112926645B (en) | Electricity stealing detection method based on edge calculation | |
Weihong et al. | Optimization of BP neural network classifier using genetic algorithm | |
CN111914082A (en) | Online knowledge aggregation method based on SOM neural network algorithm | |
CN116244484B (en) | Federal cross-modal retrieval method and system for unbalanced data | |
Skorpil et al. | Back-propagation and k-means algorithms comparison | |
CN107798331A (en) | From zoom image sequence characteristic extracting method and device | |
CN117056763A (en) | Community discovery method based on variogram embedding | |
CN112435034A (en) | Marketing arbitrage black product identification method based on multi-network graph aggregation | |
Hu et al. | Tree species identification based on the fusion of multiple deep learning models transfer learning | |
CN112015854A (en) | Heterogeneous data attribute association algorithm based on self-organizing mapping neural network | |
Zhang | Granular Neural Networks | |
CN111598119A (en) | Image clustering method based on residual error network | |
Anke et al. | Clustering analysis of gene data based on PCA and SOM neural networks | |
Zhang et al. | Multi-user Power Load Forecasting Based on K-means and Deep Neural Network |
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 |