An improved approach for association rule extraction using fuzzy formal concept analysis /
Ebtesam Elhossiny Elhossiny Hassan Shemis
An improved approach for association rule extraction using fuzzy formal concept analysis / أسلوب مُحسُّن لاستخلاص قواعد الارتباط باستخدام تحليل المفهوم الصوري الفازي Ebtesam Elhossiny Elhossiny Hassan Shemis ; Supervised Hesham Ahmed Hefny , Ahmed Mohammed Gadallah - Cairo : Ebtesam Elhossiny Elhossiny Hassan Shemis , 2018 - 199 Leaves : charts ; 30cm
Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Computer and Information Science
Intuitively, extensive involvement of the computerized system in almost all life aspects produces a massive amount of data. Such data needs to be processed, analyzed, retrieved and mined. Formal concept analysis (FCA) is a powerful tool for handling data manipulation tasks (retrieval, analysis, and mining). However, the classical FCA can only handle binary data directly. Therefore, it handles quantitative data by mapping them to binary values through dividing attribute range into a set of disjoint intervals. In consequence, FCA suffers from crisp boundaries problem regarding quantitative datasets. One of the most promising ways to overcome such deficiency in FCA is the adoption of the fuzzy set and fuzzy logic theory. In this case, fuzzy FCA (FFCA) can easily deal with such clear-cut boundaries in quantitative datasets. This dissertation concerned mainly with the ability to perform data mining in large data sets with the aid of FFCA. The proposed approach introduces the notion of one-sided fuzzy iceberg lattice which accelerates the entire mining process and best suits the association rule mining approach. The proposed fuzzy iceberg lattice contains all frequent closed itemsets associated with their corresponding fuzzy supports. Hence, it presents a straight-forward way for extracting association rules. Furthermore, in this dissertation, two enhanced algorithms, object-based and attribute-based, for extracting fuzzy concepts using FFCA are proposed
Formal concept analysis (FCA) Fuzzy FCA (FFCA) Mining
An improved approach for association rule extraction using fuzzy formal concept analysis / أسلوب مُحسُّن لاستخلاص قواعد الارتباط باستخدام تحليل المفهوم الصوري الفازي Ebtesam Elhossiny Elhossiny Hassan Shemis ; Supervised Hesham Ahmed Hefny , Ahmed Mohammed Gadallah - Cairo : Ebtesam Elhossiny Elhossiny Hassan Shemis , 2018 - 199 Leaves : charts ; 30cm
Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research - Department of Computer and Information Science
Intuitively, extensive involvement of the computerized system in almost all life aspects produces a massive amount of data. Such data needs to be processed, analyzed, retrieved and mined. Formal concept analysis (FCA) is a powerful tool for handling data manipulation tasks (retrieval, analysis, and mining). However, the classical FCA can only handle binary data directly. Therefore, it handles quantitative data by mapping them to binary values through dividing attribute range into a set of disjoint intervals. In consequence, FCA suffers from crisp boundaries problem regarding quantitative datasets. One of the most promising ways to overcome such deficiency in FCA is the adoption of the fuzzy set and fuzzy logic theory. In this case, fuzzy FCA (FFCA) can easily deal with such clear-cut boundaries in quantitative datasets. This dissertation concerned mainly with the ability to perform data mining in large data sets with the aid of FFCA. The proposed approach introduces the notion of one-sided fuzzy iceberg lattice which accelerates the entire mining process and best suits the association rule mining approach. The proposed fuzzy iceberg lattice contains all frequent closed itemsets associated with their corresponding fuzzy supports. Hence, it presents a straight-forward way for extracting association rules. Furthermore, in this dissertation, two enhanced algorithms, object-based and attribute-based, for extracting fuzzy concepts using FFCA are proposed
Formal concept analysis (FCA) Fuzzy FCA (FFCA) Mining