Authors : N.S. Nithya and K. DuraiSwamy
Abstract: Fuzzy Association Rule Mining algorithm is very efficient for diagnosis, prognosis and treatment of diseases in medical field compared to other classification technique. But it suffers from exponential growth of rules produced. Identifying the most important risk factor is one of the main tasks in medical data mining. To obtain these objectives the new algorithm using information gain ranking based weight for fuzzy associative classification is proposed. The ranking of attributes eliminates irrelevant attributes and assign weight value used for assessing the risk factor of diseases. Elimination of irrelevant attributes and ranking used to extract important rules in fuzzy association rule mining which reduce the computation time and increase the classification accuracy. The results are verified using the breast cancer dataset, heart diseases dataset with different categories of attributes to demonstrate the effectiveness of the proposed approach.
N.S. Nithya and K. DuraiSwamy, 2014. Assessing the Medical Data using Ranking Based Weighted Fuzzy Associative Classifier. International Journal of Soft Computing, 9: 20-26.