Dissertation > Excellent graduate degree dissertation topics show

Data Mining fuzzy system

Author: YangCan
Tutor: ZhuShanAn
School: Zhejiang University
Course: Control Theory and Control Engineering
Keywords: Data Mining Fuzzy Systems Universal Approximation Rulebase Fuzzy Modeling Two methods System implementation Exponential growth Mathematical structure Characterized Contact
CLC: TP18
Type: Master's thesis
Year: 2005
Downloads: 187
Quote: 1
Read: Download Dissertation

Abstract


The goal of data mining is to discover some important patterns and trends, from the complex data and really understand the meaning of the data, the fuzzy system is the \The fuzzy system is very flexible mathematical structure, is an efficient universal approximator. More importantly, many hidden in the data of the knowledge of the rule base of the fuzzy system for us, and the form of rules used by people. However, this does not mean that the fuzzy system has been flawless. 1. Many of the existing fuzzy modeling method is only from the function approximation to study this angle, that is, how to improve the speed of function the approximation precision as well as modeling. This is just from the fuzzy system the universal nature approximation fuzzy modeling, how to more fully take advantage of various information, especially data from it? 2 when faced with high-dimensional data, the fuzzy system is still facing forward to the curse of dimensionality: the exponential growth of the fuzzy rules; large number of parameters to be fit but only very sparse data scattered in the high-dimensional space. In this paper, these two problems, our own methods. We are dealing with the rules of explosion basic idea comes from clustering. The so-called \So that the number of rules will not be with the growth of the number of dimensions and the exponential growth, but linked with the characteristics of the data itself. 2 in order to obtain more effective domain of division, we study the existing clustering methods, compare their advantages and disadvantages, and the final two fuzzy modeling method based on the MCV clustering. This paper discusses the characteristics of these two methods to obtain the membership function, as well as the nature of the parameter estimation, and compared with other classical methods of these two methods. From data mining perspective, our method not only has good predictive ability, but also provide a more concise rule base, and provide a better tool for the analysis of the nature of the problem and found that the data implied knowledge. How to deal with high dimensional problems? Input selection is a critical first step. It not only plays a role of dimensionality reduction, but also increase the interpretability of the model and to reduce the amount of computation. In this paper, we propose two input selection method based on common sense, one is based on the sensitivity analysis of the input selection, the other is based on the consistency of data input options. The starting point of these two methods seems to be completely different, in fact, is equivalent, our analysis suggests that the contact on the nature of these two methods, a large number of

Related Dissertations

  1. Research on Learning-Based Super-Resolution and SR System Implementation,TP391.41
  2. Improvement and Implementation of Lean Production Management System in Multi-varieties and Small-batch,TH186
  3. Design and implementation of the management system based on the the STRUTS framework of the contract,TP311.52
  4. Research on Special Funds of Information Supervision System,F812.2
  5. Researched the Construction of the Corporate Social Responsibility Accounting in China,F235
  6. PLN 's enterprise resource planning system construction plan,F272
  7. The Research of Human Dignity Problem in the Design and Implementation of Social Security System,D632.1
  8. The Shanghai of Xingping business management system software,TP311.52
  9. The Management System and Implementation of Prison System of Song Dynasty,D926.7
  10. Research and Implementation of multiple evaluation in the C programming course,TP312.1-4
  11. The Design and Implementation of CRM customer relationship management system,TP311.52
  12. The Design and Implementation of the system of the the ABC computer-aided audit,TP311.52
  13. Southwest ATMB investment project management system design and implementation,TP311.52
  14. Development Planning on Digital Newspapering of Shenyang Daily Newspapering Group,G219.2-F
  15. The Study on Traditional Rural Cooperative Medical System in Fujian Province,R197.1
  16. Design and Implementation of Commercial Bank’s Core Business System,F832.2
  17. Study on Effect of Implemented Evaluation of Science and Technology Innovation Policy of Liaoning Province,G322.7
  18. Design and Implementation of the college campus card system,TP311.52
  19. Based collaborative service supply chain logistics management system , research,F274
  20. Supply enterprise materials management system design and implementation,TP311.52

CLC: > Industrial Technology > Automation technology,computer technology > Automated basic theory > Artificial intelligence theory
© 2012 www.DissertationTopic.Net  Mobile