Rough and Soft Set Approaches for Attributes Selection of Traditional Malay Musical Instrument Sounds Classification

Feature selection or attribute reduction is performed mainly to avoid the 'curse of dimensionality' in the large database problem including musical instrument sound classification. This problem deals with the irrelevant and redundant features. Rough set theory and soft set theory proposed by Pawlak and Molodtsov, respectively, are mathematical tools for dealing with the uncertain and imprecision data. Rough and soft set-based dimensionality reduction can be considered as machine learning approaches for feature selection. In this paper, the authors applied these approaches for data cleansing and feature selection technique of Traditional Malay musical instrument sound classification. The data cleansing technique is developed based on matrices computation of multi-soft sets while feature selection using maximum attributes dependency based on rough set theory. The modeling process comprises eight phases: data acquisition, sound editing, data representation, feature extraction, data discretization, data cleansing, feature selection, and feature validation via classification. The results show that the highest classification accuracy of 99.82% was achieved from the best 17 features with 1-NN classifier.