1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Title:	A fuzzy gravitational search algorithm for association rule mining
Authors:	Fatemeh Alikhademi (P65676)
Supervisor:	Suhaila Zainudin, Dr.
Keywords:	Rule mining Data mining Fuzzy rules Algorithm Dissertations, Academic -- Malaysia
Issue Date:	4-Aug-2015
Description:	Association Rule Mining (ARM) is a data mining task that extracts relations between items based on the items’ frequency. ARM regards items with high frequency as more interesting than items with low frequency. In quantitative datasets, each item will be separated into a large range of values. Therefore, items with low frequencies may not be considered as interesting. Hence, the possibility of extracting potentially interesting relations between these items will be decreased. Many methods were proposed to solve this issue, such as the Sharp Boundary discretization method that groups each item into intervals with crisp boundaries which do not overlap. However, associations in data mining do not necessarily have crisp boundaries and an item may be associated with more than one interval which contains an overlap, otherwise known as a fuzzy boundary. Therefore, a combination of Fuzzy Logic (FL) and Gravitational Search Algorithm (GSA) is proposed in this research. FL will group items into overlapping intervals and fuzzy rules will be generated from the interesting items. In order to enhance the quality of the generated fuzzy rules, a combination of S and Z fuzzy shapes are combined with GSA to generate the appropriate membership functions for each item. The performances of the proposed methods were evaluated over 10 KEEL and Bilkent datasets with different sizes. The results are compared with the results of ARM using the conventional Apriori algorithm and Evolutionary methods. The results show the efficiency of the proposed method to extract more accurate rules with an average accuracy close to 96% for small datasets, 91.1% for medium datasets and 78.1 % for large datasets.,Master of Artificial Intelligence
Pages:	147
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

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