Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/230610
Title: Investigating Soft Computing Techniques To Design And Implement Algorithms To Extract Useful Patterns From Large Datasets
Researcher: Gagnani, Lokesh Pitambar
Guide(s): Wandra, Kalpesh H.
Keywords: Engineering and Technology,Computer Science,Computer Science Artificial Intelligence
University: C.U. Shah University
Completed Date: 2018
Abstract: Due to the widespread use of data nowadays in the Internet era there is an extreme newlineneed to organize these large amount of data as well as it useful extraction for newlineanalysis. The IT industry, especially the multinational companies, medical newlineresearch organization around the world is facing the problem of data increasing in newlinelarge amounts on day-to-day basis. Hence there is extensive requirement to newlineanalyze these data and obtain meaningful and useful data. Data Mining has come newlineinto existence for extraction of useful patterns from these large data. However the newlinetraditional algorithms or methods are not efficient for it. Soft Computing has newlineemerged as a hot research topic in this extraction. newlineSoft Computing aka Computational Intelligence are the newest method for newlineoptimization of data mining tasks. The data mining tasks include clustering, newlineclassification and association rule mining. Further soft computing tools itself are newlinevery suitable for solving the problems of data mining because its characteristics of newlinegood robustness, self-organizing adaptive, parallel processing, distributed storage newlineand high degree of fault tolerance. Soft Computing encompasses the Swarm newlineIntelligence, Machine Learning, Fuzzy Logic, etc. other methods. The Swarm newlineIntelligence techniques are compared based on benchmark functions and best one newlineis taken into consideration for further data mining tasks. The Clustering task is newlineoptimized by the hybridization of Kernel-based FCM (KFCM), Particle Swarm newlineOptimization (PSO) and Intelligent Firefly Algorithm (IFA). The Classification newlinetask is optimized by the hybridization of Support Vector Machine (SVM). newlineSimplified Swarm Optimization (SSO-ELS) and Particle Swarm Optimization newline(PSO). The Multi Objective Association Rule (MOPAR) results are optimized by newlinethe Particle Swarm Optimization (PSO). The results obtained are compared with newlineexisting methods and performance is better. newline
Pagination: xvi, 125
URI: http://hdl.handle.net/10603/230610
Appears in Departments:Department of Computer Engineering

Files in This Item:
File Description SizeFormat 
10. chapter 5.pdfAttached File286.18 kBAdobe PDFView/Open
11. chapter 6.pdf12.83 kBAdobe PDFView/Open
12. chapter 7.pdf83.57 kBAdobe PDFView/Open
13. chapter 8.pdf15.78 kBAdobe PDFView/Open
3. title.pdf24.6 kBAdobe PDFView/Open
4. certificate.pdf900.53 kBAdobe PDFView/Open
5. preliminary pages(final).pdf128.83 kBAdobe PDFView/Open
6. chapter 1.pdf277.9 kBAdobe PDFView/Open
7. chapter 2.pdf213.48 kBAdobe PDFView/Open
8. chapter 3.pdf698.77 kBAdobe PDFView/Open
9. chapter 4.pdf24.91 kBAdobe PDFView/Open


Items in Shodhganga are protected by copyright, with all rights reserved, unless otherwise indicated.