An Independent Rough Set Approach Hybrid with Artificial Bee Colony Algorithm for Dimensionality Reduction | Science Publications

American Journal of Applied Sciences

Keppana G. Thanushkodi¹  Nambiraj Suguna¹ 

关键词: Artificial bee colony;    rough set;    k-nearest neighbor;    genetic algorithm;    dimensionality reduction;    proposed approach;    classification accuracy;    data mining;    Bees Colony Optimization (BCO);    attribute reduction;   

DOI  :  10.3844/ajassp.2011.261.266

学科分类：自然科学（综合）

来源: Science Publications

【 摘 要 】

Problem statement: Dimensionality reduction is viewed as an important pre-processingstep for pattern recognition and data mining. As the classical rough set model considers the entireattribute set as a whole to find the subset, comparing all possible combinations of sets of attributes isdifficult. Approach: In this study, we have introduced an improved Rough Set-based AttributeReduction (RSAR) namely Independent RSAR hybrid with Artificial Bee Colony (ABC) algorithm,which finds the subset of attributes independently based on decision attributes (classes) at first andthen finds the final reduct. Initially the instances are grouped based on decision attributes. Then theQuick Reduct algorithm is applied to find the reduced feature set for each class. To this set of reducts,the ABC algorithm is applied to select a random number of attributes from each set, based on theRSAR model, to find the final subset of attributes. Results: The performance is analyzed with fivedifferent medical datasets namely Dermatology, Cleveland Heart, HIV, Lung Cancer and Wisconsinand compared with six other reduct algorithms. The reduct from the proposed approach reaches greateraccuracy of 92.36, 86.54, 86.29, 83.03 and 88.70 % respectively. Conclusion: The experiments statesthat the proposed approach reduces the computational cost and improves the classification accuracywhen compared to some classical techniques.

【 授权许可】

Unknown