Dissertation > Excellent graduate degree dissertation topics show
Research on Directional Clustering and It’s Applications
Author: XiuYu
Tutor: WangShiTong
School: Jiangnan University
Course: Applied Computer Technology
Keywords: Cluster analysis High-dimensional data Directional data Directional metrics Gene expression data Text Clustering Principle of maximum entropy
CLC: TP18
Type: Master's thesis
Year: 2006
Downloads: 165
Quote: 0
Read: Download Dissertation
Abstract
|
Cluster analysis is the use of mathematical methods and processing of a given object classification , its aim is to discover the hidden distribution structure in the data set . Over the last decade , greatly improve the ability to access and production data . By clustering analysis algorithm found these hidden data within the data distribution model, as well as valuable linkages exist between the data attribute . The clustering techniques widely used in bioinformatics, archeology, psychology, computer image processing, information retrieval, engineering controls , and many other areas . Often encountered in the practical application of high -dimensional data , such as gene expression data, text data, multimedia data . Due to the universality of this data , the clustering analysis of high-dimensional data is of great significance . In this paper, the directivity and Cluster Analysis of high-dimensional data problems . Including gene expression data clustering analysis and text clustering analysis, and proposed some solutions to the problem , has a certain theoretical and practical significance . The main work of this paper are: 1 ) analysis of high-dimensional data directional characteristics , gene expression data , and on this basis to construct a new similarity measure - directional similarity suitable for gene expression clustering algorithm DSCM . The algorithm overcomes the initialization sensitivity the other directivity clustering algorithm when clustering of gene expression , to a certain extent can automatically determine the number of classes as well as having a certain outlier detection function , and thus has certain robustness . A large number of simulation results show that the algorithm has good performance . 2 ) the introduction of the principle of maximum entropy in information theory to the classic spherical text clustering , maximum entropy objective function and maximum entropy for text clustering spherical text clustering algorithm is constructed . The new algorithm can avoid local minima and the global minimum , has initially solved the classic spherical text clustering algorithm initialization sensitivity problem , and the clustering performance improved .
|
Related Dissertations
- Research and Implementation of Mining Implicit User Interest,TP311.13
- Development of EST-SSR Primers and Application in Analysis of Genetic Realtionships in Tree Peony,S685.11
- Pre-hypertension syndrome characteristics,R259
- Comprehensive Quality Assessment of College Students,G645.5
- Studieson Effects of Soybean Species on Yuba and Initial Establishment of Quality Evalution System for Yuba,TS214.2
- ISSR Analysis of Genetic Diversity on 21 Lotus(Nelumbo Nucifera) Cultivars,S682.32
- Research on the Soil Environmental Function Zoning,X321
- Comparison of Gene Expression Data Cluster Methods and Gene Network Construction for Phytophthora Sojae Genes,S435.651
- Study on Heterosis and Genetic Basis of Soybean,S565.1
- Evaluation on Forage Quality and Biomass Energy Characters of Inbred Vegetative Lines of Napier Grass,S543.9
- The Design and Implementation of Bicluster Data Analyzing Software,TP311.52
- Clustering Method Research Based on Divided and Conquered Method,TP311.13
- Elements of Thirty Kinds of Proprietary Chinese Medicines and Classification Based on Elements,R286.0
- Design to E-learning System in Senior Vocational School Base on Moodle,TP311.52
- Research of Text Clustering on Food Complaint Documents Based on Ontology,TP391.1
- The Research of Prairie Road Light Environment Effects on Physiological Indicators of Drivers,U491.254
- Improved Mathematical Model of Spatial Information Processing Based on Fuzzy Technology,O159
- Based embedded software fault-tolerant data flow anomaly detection,TP368.1
- Research on Classification of Colleges in Our Country Based on Clustering Technology of Data Mining,TP311.13
- Research and Application on Short Message Text Clustering,TP391.1
- Research on the Platform for Clustering Graph Data and Its Implementation,TP311.13
CLC: > Industrial Technology > Automation technology,computer technology > Automated basic theory > Artificial intelligence theory
© 2012 www.DissertationTopic.Net Mobile
|