Frequent pattern mining is used to derive association rules. Association rules specify relativity of target class with rest of the feature(s). The Apriori and FP-growth algorithms are the most famous algorithms used for frequent pattern mining. Classification with feature selection approach is also widely used. This paper provides a detailed study of frequent pattern mining using BitApriori algorithm and use mined association rules for performance improvement of unbalanced data classification. We present a model called FPCM which first mine association rules. Mined association rules are than used for features selection. In final phase, selected features are used in unbalanced data classification using decision tree classifier. Our model shows improved accuracy as compare to the past studies.

Frequent pattern mining, Apriori, BitApriori, Unbalanced data classification, machine learning

[1] Varsha Mashoria, Anju Singh, "Literature Survey on Various Frequent Pattern Mining Algorithm", IOSRJEN, Vol-3, Jan-2013.
[2] Sumit Aggarwal, V Singal, "A Survey on Frequent Pattern Mining Algorithms", (IJERT) ISSN: 2278-0181 4, April – 2014
[3] Le Thi Thanh Nhan, Thi T T Nguyen, Tae Chong Chung, "BitApriori: An Apriori-Based Frequent Itemsets Mining Using Bit Streams", IEEE, 2010
[4] E. Ansari, G.H. Dastghaibifard, M. Keshtkaran, H.Kaabi, "Distributed Frequent Itemset Mining using Trie Data Structure", IAENG Inter. Journal of Comp. Sci., 2008
[5] Charu C. Aggarwal, Mansurul A. Bhuiyan and Mohammad Al Hasan, "Frequent Pattern Mining Algorithms: A Survey", Switzerland, Springer International Publishing Switzerland, 2014
[6] Abdul Rahaman Wahab Sait, and Dr.T.Meyappan, "Data preprocessing and Transformation Technique to Generate Pattern From the WebLog", Dubai (UAE), ICSIS’2014, Oct 17-18, 2014.
[7] Abdolrashid Rezvani, J Hosseinkhani, "Enhancing the Performance of BitApriori Algorithm in Data Mining using an Effective Data Structure", Switzerland, IJACSIT,Vol. 4, No. 3, 2015, Page: 85-92.
[8] Samiksha Kankane, V garg, "A survey paper on : Frequent Pattern Analysis Algorithm from the Web Log Data", IJCA, Vol-119, June-2015.
[9] D.Usha, Dr.K.Rameshkumar, "A Complete Survey on application of Frequent Pattern Mining and Association Rule Mining on Crime Pattern Mining", International Journal of Advances in Computer Science and Technology, vol-3, April-2014.
[10] Astha Agrawal, Herna L Victor, Eric Paquet, “SCUT: Multi-Class Imbalanced Data Classification using SMOTE and Cluster-based Undersampling”, SCITEPRESS, 2015.
[11] Isaac Triguero, Sara D Rio, V Lopez, J Bacardit, J M Benitez, F Herrera, “ROSEFW-RF: The winner algorithm for the ECBDL’14 big data competition: An extremely imbalanced big data bioinformatics problem”, Knowledge Based System, Elsevier, 28 May 2015.
[12] Sotiris Kotsiantis, Dimitris Kanellopoulos, Panayiotis Pintelas, “Handling imbalanced datasets: A review”, GESTS International Transactions on Computer Science and Engineering, Vol.30, 2006.
[13] Chang Wan, “Test-Cost Sensitive Classification on Data with Missing Values in the Limited Time”, Springer-Verlag Berlin Heidelberg 2010.
[14] K. Rajeswari, “Feature Selection by Mining Optimized Association Rules based on Apriori Algorithm”, IJCA, Vol. 119, Jun-2015
[15] Guangtao Wang, Qinbao Song, “Selecting Feature Subset via Constraint Association Rules”, Springer-Verlag Berlin Heidelberg 2012
[16] UCI repository, http://tunedit.org/repo/UCI/heart-statlog.arff.