Study of Neural Network Classing Model Based on Clustering

ZHU GUOQIANG,LIU HOUQUAN

2008-01-01

Abstract:Cases-choice in tax declaration is an interesting topic.An approach of cases-choice in tax declaration,Neural Network Classing Model Based on Clustering,is proposed.The proposed method divides samples into different sub-classes,and then builds neural network classing model.The experiments show that the proposed method is efficient.

What problem does this paper attempt to address?

Fuzzy Naive Bayes Classifier Based on Fuzzy Clustering

Yongchuan Tang,Wuming Pan,Haiming Li,Yang Xu

DOI: https://doi.org/10.1109/icsmc.2002.1176401

2002-01-01

Abstract:Despite its unrealistic independence assumption, the Naive Bayes classifier is remarkably successful in practice. In the Naive Bayes classifier, all variables are assumed to be nominal variables, it means that each variable has a finite number of values. But in large databases, the variables (or fields) often take continuous values or have a large number of numerical values. So many researchers discussed the discretization (or crisp partitioning) of the domain of the continuous variables. We generalize the Naive Bayes classifier to the situation in which the fuzzy partition of the variable domains instead of discretization is taken. Therefore each variable in the Fuzzy Naive Bayes classifier can take a linguistic value or fuzzy set. From the observed data set one method of estimating the conditional probabilities in the Fuzzy Naive Bayes classifier is proposed in this paper. For each numeric input the method to predict its class label using the fuzzy Naive Bayes classifier is presented. In the training phase of the classifier, the training data (just including the feature variables without class labels) is first clustered in an unsupervised way by fuzzy c-means or a similar algorithm. Then the optimal cluster centers of training data are used to determine the fuzzy partition of the feature variables space. This generalization can decrease the complexity of learning optimal discretization which the classical Naive Bayes Classifier often faces, reduce the loss of information because of the discretization and increase the power of dealing with imprecise data and the large databases. Some well-known classification problems in the machine learning field have been tested in this paper, the results show that the Fuzzy Naive Bayes classifier is an alternative and effective tool to deal with the classification problem which has continuous variables.
[A Study on Clustering Analysis of Arrhythmias].

Zetao Lin,Yaozheng Ge,Guoliang Tao

DOI: https://doi.org/10.3321/j.issn:1001-5515.2006.05.018

2006-01-01

Journal of Biomedical Engineering

Abstract:According to the characteristics of ECG analysis, large data quantum, high accuracy demand and real-time, a classified algorithm of arrhythmia based on clustering analysis is presented in this paper. According to "things-of-one-kind-come-together" principle, this algorithm uses the similarities of cases with same kind of heart disease at the same time, includes the factors of the individual difference to analyze arrhythmias by clustering QRS complex waveform and rhythm analysis as the subordinate method. Verified by eight records of MIT-BIR standard heart electricity database, the probability of correct clustering reaches above 90%, which shows that this algorithm can analyze arrhythmias effectively.
A Noise-Resistant Clustering Algorithm for Switching Regression Models

杨小兵,何灵敏,孔繁胜

DOI: https://doi.org/10.3969/j.issn.1673-4785.2009.06.005

2009-01-01

Abstract:Clustering methods for switching regression models usually neglect the effects of noise.As a result,errors usually exist if clustering is carried out in a noisy environment.In order to overcome the effects of noise,a new clustering algorithm,a noise-resistant clustering algorithm,was proposed.The algorithm partitions the dataset into two sub-datasets,a noiseless dataset and a noisy dataset,and then performs clustering analysis on the noiseless dataset to estimate parameters.By continuous simultaneous adjustment of the noisy and noiseless datasets and by continuously revising estimated parameters,the results of clustering were improved.Simulation experiments demonstrated that the algorithm efficiently clusters noisy datasets and can provide good estimates of parameters.
Study Of Artificial Neural Network Model Based On Fuzzy Clustering

Gang Wang,Lihua Huang,Chenghong Zhang

DOI: https://doi.org/10.1109/WCICA.2006.1712857

2006-01-01

Abstract:During the application of artificial neural network, there is a question of how to deal with a great lot of the complicated sample set and the long time of training. The article proposes an artificial neural network model based on fuzzy clustering. Firstly, it clusters the training data into different sub-data using by Fuzzy Clustering Model. Subsequently, it uses different artificial neural network to train the sub-data. Because of doing this, the number and complexity of training data by every artificial neural network is reduced and the efficiency of every artificial neural network is enhanced greatly. Lastly, the article uses the UCI's databases to prove the utility of the new artificial neural network and gets the satisfied answers.
Evaluation Model on Organizational Networking Based on Artificial Neural Networks

ZHU Qi-hong,ZHANG Gang

DOI: https://doi.org/10.3969/j.issn.1001-3695.2007.06.078

2007-01-01

Abstract:By studying given sample modes,a three-layer neural network mode could obtain experience,knowledge,subjective judgment,and the ability to weigh the importance of targets from evaluation specialists,so as to ensure objectivity of evaluation results.Examples show that neural networks are feasible and efficient for the evaluation of the organizational networking.
Application of Neural Network Based on Fuzzy Clustering in the Classification Algorithm of Data Mining

WANG Gang,HUANG Lihua,ZHANG Chenghong

DOI: https://doi.org/10.3321/j.issn:1000-7857.2007.15.013

2007-01-01

Abstract:The common classification algorithms, such as Decision Tree, Bayesian classification and Artificial Neural Network, provide an important foundation for the classification algorithm of data mining. But due to huge amounts of data, these algorithms face some challenges in time efficiency, robustness and accuracy. This paper proposes an artificial neural network model based on fuzzy clustering. First, it clusters the training data into different sub-data using the Fuzzy Clustering Model. Subsequently, it uses various artificial neural networks to train the sub-data. After that, the number and complexity of training data is reduced and the efficiency of the artificial neural network is enhanced greatly. In the end, the UCI's databases are used to prove the usefulness of the new classification algorithm. The results show the validity of application of the Neural Network based on fuzzy clustering in the classification of data mining.
Diagnosis-Guided Deep Subspace Clustering Association Study for Pathogenetic Markers Identification of Alzheimer's Disease Based on Comparative Atlases

Cui-Na Jiao,Junliang Shang,Feng Li,Xinchun Cui,Yan-Li Wang,Ying-Lian Gao,Jin-Xing Liu

DOI: https://doi.org/10.1109/JBHI.2024.3372294

Abstract:The roles of brain region activities and genotypic functions in the pathogenesis of Alzheimer's disease (AD) remain unclear. Meanwhile, current imaging genetics methods are difficult to identify potential pathogenetic markers by correlation analysis between brain network and genetic variation. To discover disease-related brain connectome from the specific brain structure and the fine-grained level, based on the Automated Anatomical Labeling (AAL) and human Brainnetome atlases, the functional brain network is first constructed for each subject. Specifically, the upper triangle elements of the functional connectivity matrix are extracted as connectivity features. The clustering coefficient and the average weighted node degree are developed to assess the significance of every brain area. Since the constructed brain network and genetic data are characterized by non-linearity, high-dimensionality, and few subjects, the deep subspace clustering algorithm is proposed to reconstruct the original data. Our multilayer neural network helps capture the non-linear manifolds, and subspace clustering learns pairwise affinities between samples. Moreover, most approaches in neuroimaging genetics are unsupervised learning, neglecting the diagnostic information related to diseases. We presented a label constraint with diagnostic status to instruct the imaging genetics correlation analysis. To this end, a diagnosis-guided deep subspace clustering association (DDSCA) method is developed to discover brain connectome and risk genetic factors by integrating genotypes with functional network phenotypes. Extensive experiments prove that DDSCA achieves superior performance to most association methods and effectively selects disease-relevant genetic markers and brain connectome at the coarse-grained and fine-grained levels.
Three-way Decisions Based Bayesian Network Classifier

Lu Wang,Xiuyi Jia,Yannan Gu

DOI: https://doi.org/10.13232/j.cnki.jnju.2016.05.009

2016-01-01

Abstract:The traditional classification models are always two‐way decision models ,which are based on the attributes or features of objects to accept or reject objects .However ,in many cases ,because of the inaccurate information ,a number of objects belonging to different classes have the similar probability value .Therefore ,classical two‐way decisions based classifiers usually have a high probability to misclassify the vague objects .To obtain a lower misclas‐sification rate ,three‐way decisions based classifiers adopt the deferment decision to deal with those vague objects . Those vague objects are classified into boundary region waiting for further processing .In order to facilitate the calcu‐lation ,most three‐way decision classifications are based on the single evaluation function .In single function based three‐way decisions ,all objects will be classified into the corresponding regions by comparing their conditional probabilities with the decision thresholds .The decision thresholds can be computed by three‐way decision‐theoretic rough set model ,while the conditional probability of an object has to be provided by a specific classifier .In the com‐putational efficiency ,Naive Bayesian Rough set model combines the Naive Bayes and the three‐way decision , resulting high computational efficiency .However ,Naive Bayesian model is based on the assumption of conditional in‐dependence so that the obtained values are usually inaccurate .In view of the problems existing in the Naive Bayesian Rough set model ,this paper proposes a three‐way decisions based Bayesian network classifier ,which introduces the Bayesian network model and combines the computation of the object’s conditional probability and decision thresholds together .Experimental results show that our proposed classifier can have a high accuracy .
Model Matching: A Novel Framework to Use Clustering Strategy to Solve the Classification Problem.

Zhiyi Duan,Limin Wang,Minghui Sun

DOI: https://doi.org/10.1109/access.2019.2922199

IF: 3.9

2019-01-01

IEEE Access

Abstract:It is a common practice to handle labeled data with classifiers and unlabeled ones with clusterings. The traditional Bayesian network classifiers (BNC$^{\mathcal {T}}\text{s}$ ) learned from labeled training set $\mathcal {T}$ directly map the unlabeled test instance into the network structure to calculate the conditional probability for the classification, which neglects the information hidden in the unlabeled data and will result in classification bias. To address this issue, we propose a novel learning framework, called model matching, that uses the “clustering” strategy to solve the classification problem. The labeled data is divided into several clusters according to the different class label to learn a set of BNC$^{\mathcal {T}}\text{s}$ and a corresponding set of BNC$^{p}\text{s}$ is built for each unlabeled test instance. To make a classification, the cross entropy method is applied to compare the structural similarity between BNC$^{\mathcal {T}}$ and BNCp. The extensive experimental results on 46 datasets from the University of California at Irvine (UCI) machine learning repository demonstrate that for BNCs model matching helps improve the generalization performance and outperforms the several state-of-the-art classifiers like tree-augmented naive Bayes and Random forest.
Transformer-Based Hierarchical Clustering for Brain Network Analysis

Wei Dai,Hejie Cui,Xuan Kan,Ying Guo,Sanne van Rooij,Carl Yang

DOI: https://doi.org/10.48550/arXiv.2305.04142

IF: 5.414

2023-05-06

Machine Learning

Abstract:Brain networks, graphical models such as those constructed from MRI, have been widely used in pathological prediction and analysis of brain functions. Within the complex brain system, differences in neuronal connection strengths parcellate the brain into various functional modules (network communities), which are critical for brain analysis. However, identifying such communities within the brain has been a nontrivial issue due to the complexity of neuronal interactions. In this work, we propose a novel interpretable transformer-based model for joint hierarchical cluster identification and brain network classification. Extensive experimental results on real-world brain network datasets show that with the help of hierarchical clustering, the model achieves increased accuracy and reduced runtime complexity while providing plausible insight into the functional organization of brain regions. The implementation is available at https://github.com/DDVD233/THC.
Module partitioning for multilayer brain functional network using weighted clustering ensemble

Zhuqing Jiao,Xuelian Ming,Yin Cao,Chun Cheng,Shui-Hua Wang

DOI: https://doi.org/10.1007/s12652-019-01535-4

IF: 3.662

2019-10-09

Journal of Ambient Intelligence and Humanized Computing

Abstract:Module can not only affect the integration of network functions, but also contribute to understand the characteristics of local connections in the network. However, the connection between nodes in the network changes with the passage of time, and the module structure changes accordingly. To overcome this drawback, we propose a method of applying weighted clustering ensemble to partition multilayer brain functional networks into modules. Firstly, k-means clustering is adopted to carry out base clustering for a certain layer of functional network for several times, and each clustering is corresponding to a subordinate matrix and a similarity matrix. Then clustering validity index is used to assess each partitioning and the assessed values are taken as the weights of similarity matrix. Finally, the weighted similarity matrix is partitioned by means of fuzzy C-means clustering, and the results are evaluated by modularity function to obtain the optimal partitioned modules. Experimental results show that the effect of module partitioning resulting from weighted clustering ensemble is better than that of other comparable methods. The proposed framework could be promising to analyze the differences between corresponding modules of patients with Alzheimer’s disease and normal people, so as to better understanding some dynamical pathological changes in brain connectome of patients.

computer science, information systems,telecommunications, artificial intelligence
Neural Networks for Clustering Analysis of Molecular Data

Lin Wang,Minghu Jiang,Xiaofang Tang,Qiuqi Ruan,Baozong Yuan,Frank Noé

DOI: https://doi.org/10.1109/ICOSP.2006.345839

2007-01-01

Abstract:In this paper hierarchical cluster and the competitive learning cluster are compared by using molecular data of large size sets. We construct a reproducible matrix to evaluate the quality of clustering, and dead nodes problem of the competitive learning network is solved by the conscience mechanism. The experimental results show that the hierarchical clustering can represent a multi-level hierarchy which show the tree relation of cluster distance, the competitive learning network has a good clustering reproducible and indicate the effectiveness of clusters for molecular data
Categorizing SHR and WKY rats by chi2 algorithm and decision tree

Ping-Rui Tsai,Kun-Huang Chen,Tzay-Ming Hong,Fu-Nien Wang,Teng-Yi Huang

DOI: https://doi.org/10.1038/s41598-021-82864-3

IF: 4.6

2021-02-10

Scientific Reports

Abstract:Abstract Classifying mental disorder is a big issue in psychology in recent years. This article focuses on offering a relation between decision tree and encoding of fMRI that can simplify the analysis of different mental disorders and has a high ROC over 0.9. Here we encode fMRI information to the power-law distribution with integer elements by the graph theory in which the network is characterized by degrees that measure the number of effective links exceeding the threshold of Pearson correlation among voxels. When the degrees are ranked from low to high, the network equation can be fit by the power-law distribution. Here we use the mentally disordered SHR and WKY rats as samples and employ decision tree from chi2 algorithm to classify different states of mental disorder. This method not only provides the decision tree and encoding, but also enables the construction of a transformation matrix that is capable of connecting different metal disorders. Although the latter attempt is still in its fancy, it may have a contribution to unraveling the mystery of psychological processes.

multidisciplinary sciences
Exploring Functional Connectivity Networks in fMRI Data Using Clustering Analysis.

Dazhong Liu,Ning Zhong,Yulin Qin

DOI: https://doi.org/10.1007/978-3-642-23605-1_17

2011-01-01

Abstract:Some approaches have been proposed for exploring functional brain connectivity networks from functional magnetic resonance imaging (fMRI) data. Based on a popular algorithm K-means and an effective clustering algorithm called Affinity Propagation (AP), a combined clustering method to explore the functional brain connectivity networks is presented. In the proposed method, K-means is used for data reduction and AP is used for clustering. Without setting the seed of ROI in advance, the proposed method is especially appropriate for the analysis of fMRI data collected with a periodic experimental paradigm. The validity of the proposed method is illustrated by experiments on a simulated dataset and a human dataset. Receiver operating characteristic (ROC) analysis was performed on the simulated dataset. Results show that this method can efficiently and robustly detect the actual functional response with typical signal changes in the aspect of noise ratio, phase and amplitude. On the human dataset, the proposed method discovered brain networks which are compatible with the findings of previous studies.
The Application of Rough Set-Neural Network Based on c-Means Clustering in Heart Disease Diagnosation

DING Bao-miao,ZHANG Yun-tao,GAO Xing-yu

DOI: https://doi.org/10.3969/j.issn.1001-9626.2007.02.026

2007-01-01

Abstract:By using c-means clustering,an algorithm is presented to discretize condition attributes in a decision table.The simplified rules are available with the rough set theory pro- cessing the discretized decision table system,and the data set obtained is then introduced to the BP network for training and testing.Based on the present model a group of data form heart diseases is processed,achieving a testing correctness percentage of 85 while the other two are 76% and 82% respectively.In the combined model of rough set and BP neural network,the tradition algorithm,i.e.the equation-distance and the equation-frequency for discretization can only have 53% and 77% of testing correctness.
A Novel Probabilistic Clustering Model for Heterogeneous Networks

Zhi-Hong Deng,Xiaoran Xu

DOI: https://doi.org/10.1007/s10994-016-5544-1

IF: 5.414

2016-01-01

Machine Learning

Abstract:Heterogeneous networks, consisting of multi-type objects coupled with various relations, are ubiquitous in the real world. Most previous work on clustering heterogeneous networks either converts them into homogeneous networks or simplifies the modeling of the heterogeneity in terms of specific objects, structures or assumptions. However, few studies consider all relevant objects and relations, and trade-off between integrating relevant objects and reducing the noises caused by relations across objects. In this paper, we propose a general probabilistic graphical model for clustering heterogeneous networks. First, we present a novel graphical representation based on our basic assumptions: different relation types produce different weight distributions to specify intra-cluster probability between two objects, and clusters are formed around cluster cores. Then, we derive an efficient algorithm called PROCESS, standing for PRObabilistic Clustering modEl for heterogeneouS networkS. PROCESS employs a balance-controlled message passing algorithm and mathematical programming for inference and estimation. Experimental results show that our approach is effective and significantly outperforms the state-of-the-art algorithms on both synthetic and real data from heterogeneous networks.
Study on the Credit Classification of Practicing Qualification Personnel in Construction Market Based on PNN

Hui Sun,Zhiqing Fan,Ying Zhou

DOI: https://doi.org/10.4028/www.scientific.net/amm.44-47.13

2010-01-01

Applied Mechanics and Materials

Abstract:Combining with the characters of the practicing qualification personnel in construction market, probabilistic neural network is brought out trying to analyze the credit classification of the practicing qualification personnel. And the impact factor of the number of neurons on the credit classification of the practicing qualification personnel is studied. Then a probabilistic neural network is built. At last, a case study is conducted by taking practicing qualification personnel as an example. The research result reveals that the method can efficiently evaluate the credit of the practicing qualification personnel, thus it could provide scientific advice to the construction enterprise to prevent relevant discreditable behaviors of some practicing qualification personnel.
Neighborhood rough set-based three-way clustering considering attribute correlations: An approach to classification of potential gout groups

Xiaoli Chu,Bingzhen Sun,Xue Li,Keyu Han,JiaQi Wu,Yan Zhang,Qingchun Huang

DOI: https://doi.org/10.1016/j.ins.2020.05.039

IF: 8.1

2020-10-01

Information Sciences

Abstract:<p>Using modern information theory to classify and identify high-risk disease groups is one of the research concerns in medical decision-making. The early diagnosis of gout is missing a single indicator, and relying on artificial labeling of disease characteristics is not only costly for decision-making, but also has a high misdiagnosis rate. Aiming at incomplete and attribute-related random large sample data, we propose a three-way clustering algorithm based on neighborhood rough sets, which is used to initially label the data, reduce the rate of misdiagnosis, and improve decision-making efficiency. Firstly, a neighborhood rough set theory in a heterogeneous information system is established. Secondly, the Best-Worst method-based neighborhood rough set attribute reduction model considering attribute correlation is constructed. Thirdly, a neighborhood rough set-based three-way clustering method for heterogeneous information system is proposed. Finally, we use 2,683 random samples and the proposed model to identify and classify potential gout patients in the samples. The results show that the proposed model can be used to mark and cluster potential gout groups in random samples without prior probability and with fuzzy decision rules, which is helpful for clinical decision-making.</p>

computer science, information systems
Recidivism early warning model based on rough sets and the improved K-prototype clustering algorithm and a back propagation neural network

Kangshun Li,Ziming Wang,Xin Yao,Jiahao Liu,Hongming Fang,Yishu Lei

DOI: https://doi.org/10.1007/s12652-021-03337-z

IF: 3.662

2021-06-12

Journal of Ambient Intelligence and Humanized Computing

Abstract:The rate of recidivism by criminals after their release from prison is high, which is harmful to society. Thus, it is socially significant to reduce their recidivism rate. This article uses public data from the state of Iowa in the United States. According to the data characteristics, such as having redundant samples and mixed attributes, we propose the following methods. First, we use a rough set attribute reduction algorithm based on probability distributions to reduce the redundant items. Second, the sample data are clustered with an improved clustering algorithm. Based on the traditional K-prototype clustering algorithm, the clustering algorithm is improved by changing the measurement method of the categorical attributes, changing the initial cluster center selection method, and weighting the attributes based on the information entropy. The clustering experiment results show that the improved clustering algorithm has a better clustering effect and higher clustering accuracy than the traditional K-prototype clustering algorithm. Finally, a back propagation neural network is used to predict the recidivism probability of the sample processed by the above algorithm. The final experimental results show that the two redundant attributes are successfully reduced by rough sets, which greatly reduces the run time of the model. Compared with the traditional K-prototype clustering algorithm, the improved K-prototype clustering algorithm proposed in this paper has a better effect on the various indicators and objective function. Finally, through neural network prediction, the prediction accuracy of this model reached 87.9%. At the same time, a large number of experiments on benchmark datasets verify the effectiveness of our proposed model.

computer science, information systems,telecommunications, artificial intelligence
A METHOD OF HIERARCHICAL DOCUMENT AUTOMATIC CLASSIFICATION IN E-RESEARCH

Yun Jian,Jiang Di,Pan Wuyun

DOI: https://doi.org/10.3969/j.issn.1000-386X.2009.11.015

2009-01-01

Abstract:Subjects crossing is very common in E-research,so it is necessary for documents in multi-subjects to be classified automatically.In accordance with high dimensions of documents in these multi-subjects,a method of hierarchical automatic classification with the thought of DC(divide and conquer) is proposed and used for E-research of comparative linguistics in E-Institutes of Shanghai Universities.First,vectors of different documents are clustered by geometric classification without matrix transpose.Then,1-dimensional feature space is formed through Fisher linear discriminant criterion.Finally,by using NBayes decision,a MCE(minimum classification error) decision,documents automatic classification is proceeded in an effective way.Experiment result indicates that the method is effective:both in close-set test and open-set test,it has good performance in precision,recall and F1.The classification process takes 0.29 sec on average.The above work offers some intelligent supports for E-research.

Study of Neural Network Classing Model Based on Clustering

Fuzzy Naive Bayes Classifier Based on Fuzzy Clustering

[A Study on Clustering Analysis of Arrhythmias].

A Noise-Resistant Clustering Algorithm for Switching Regression Models

Study Of Artificial Neural Network Model Based On Fuzzy Clustering

Evaluation Model on Organizational Networking Based on Artificial Neural Networks

Application of Neural Network Based on Fuzzy Clustering in the Classification Algorithm of Data Mining

Diagnosis-Guided Deep Subspace Clustering Association Study for Pathogenetic Markers Identification of Alzheimer's Disease Based on Comparative Atlases

Three-way Decisions Based Bayesian Network Classifier

Model Matching: A Novel Framework to Use Clustering Strategy to Solve the Classification Problem.

Transformer-Based Hierarchical Clustering for Brain Network Analysis

Module partitioning for multilayer brain functional network using weighted clustering ensemble

Neural Networks for Clustering Analysis of Molecular Data

Categorizing SHR and WKY rats by chi2 algorithm and decision tree

Exploring Functional Connectivity Networks in fMRI Data Using Clustering Analysis.

The Application of Rough Set-Neural Network Based on c-Means Clustering in Heart Disease Diagnosation

A Novel Probabilistic Clustering Model for Heterogeneous Networks

Study on the Credit Classification of Practicing Qualification Personnel in Construction Market Based on PNN

Neighborhood rough set-based three-way clustering considering attribute correlations: An approach to classification of potential gout groups

Recidivism early warning model based on rough sets and the improved K-prototype clustering algorithm and a back propagation neural network

A METHOD OF HIERARCHICAL DOCUMENT AUTOMATIC CLASSIFICATION IN E-RESEARCH