Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/542936
Title: | Improved Multiple Minimum Support Based Approaches to Mine Frequent Patterns |
Researcher: | Uday Kiran, R |
Guide(s): | Krishna Reddy, P |
Keywords: | Computer Science Computer Science Information Systems Engineering and Technology |
University: | International Institute of Information Technology, Hyderabad |
Completed Date: | 2011 |
Abstract: | The field of data mining has emerged to extract information/knowledge hidden in large newlinedatabases for better decision making. Several data mining patterns, such as association rules, newlineclustering and classification, are being proposed. Research is going on to investigate efficient newlineapproaches to extract the patterns pertaining to rarely occurring objects besides discovering newlinenew patterns. Due to its usefulness in decision making process, recently, research efforts are newlinegoing on to investigate efficient approaches for mining rare cases or patterns. Mainly, research newlineefforts are being made to investigate efficient approaches to extract rare association rules and newlinerare classes. In this thesis, we have made an effort to propose improved approaches for extracting rare association rules. newlineFrequent pattern mining is a key step in many association rule mining algorithms. In the newlinebasic model of association rules, a pattern is said to be frequent if it satisfies the user-defined newlineminimum support (minsup) threshold value. Since only a single minsup is used in the entire newlinedatabase, the basic model of frequent patterns leads to the problem known as rare item problem which is as follows: at high minsup, we miss the frequent patterns containing rare items, newlineand at low minsup, combinatorial explosion can occur, producing too many frequent patterns. newlineTo confront the rare item problem, an effort has been made in the literature to find frequent patterns with multiple minsups framework. In this framework, each item is given a constraint newlineknown as minimum item support (MIS). The notion of minimum support for a pattern is defined newlineas the minimal MIS value among all its items. Efforts are being made to propose Apriori and newline FP-growth based approaches to extract patterns under multiple minsups framework. This newlinegeneralized framework enables the user to simultaneously specify high minsup for a pattern newlinecontaining only frequent items and low minsup for a pattern containing rare items.In this thesis, we identified three opportunities for improving the |
Pagination: | 94 |
URI: | http://hdl.handle.net/10603/542936 |
Appears in Departments: | Computer Science and Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
80_recommendation.pdf | Attached File | 50.77 kB | Adobe PDF | View/Open |
abstract.pdf | 31.25 kB | Adobe PDF | View/Open | |
annexures.pdf | 62.98 kB | Adobe PDF | View/Open | |
chapter 1.pdf | 73.4 kB | Adobe PDF | View/Open | |
chapter 2.pdf | 80.08 kB | Adobe PDF | View/Open | |
chapter 3.pdf | 99.21 kB | Adobe PDF | View/Open | |
chapter 4.pdf | 72.4 kB | Adobe PDF | View/Open | |
chapter 5.pdf | 114.53 kB | Adobe PDF | View/Open | |
chapter 6.pdf | 38.48 kB | Adobe PDF | View/Open | |
content.pdf | 30.77 kB | Adobe PDF | View/Open | |
preliminary pages.pdf | 51.56 kB | Adobe PDF | View/Open | |
titlepage.pdf | 39.01 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: