IJE TRANSACTIONS C: Aspects Vol. 30, No. 9 (September 2017) 1326-1333   

PDF URL: http://www.ije.ir/Vol30/No9/C/5.pdf  
downloaded Downloaded: 43   viewed Viewed: 348

A. Moaref and V. Sattari-Naeini
( Received: January 19, 2017 – Accepted in Revised Form: July 07, 2017 )

Abstract    Feature selection for various applications has been carried out for many years in many different research areas. However, there is a trade-off between finding feature subsets with minimum length and increasing the classification accuracy. In this paper, a filter-wrapper feature selection approach based on fuzzy-rough gain ratio is proposed to tackle this problem. As a search strategy, a modified Ant Colony Optimization (ACO) algorithm is applied on filter phase. ACO has been approved to be a suitable solution in many difficult problems with graph search space such as feature selection. Choosing minimal data reductions among the subsets of features with first and second maximum accuracies is the main contribution of this work. To verify the efficiency of our approach, experiments are performed on 10 well-known UCI data sets. Analysis of the experimental results demonstrates that the proposed approach is able to satisfy two conflicting constraints of feature selection, increasing the classification accuracy as well as decreasing the length of the reduced subsets of features.


Keywords    Feature Selection, Fuzzy Rough Sets, Ant Colony Optimization, Filter-Wrapper Method


چکیده    سالهاست که مسأله‌ی انتخاب ویژگی در عرصه‌های تحقیقاتی مختلف و برای کاربردهای مختلف به کار می رود؛ درحالی‌که همواره بین پیداکردن مجموعه ویژگی‌ها با کمترین طول از یک طرف، و افزایش دقت دسته‌بند از طرف دیگر تضاد وجود دارد. در این مقاله یک روش فیلتر-دسته بند بر پایه‌ی نرخ بهره‌ی اطلاعاتی در مجموعه‌های ناهموار فازی ارائه شده است که می‌تواند از عهده‌ی این مشکل برآید. از آن­جاکه الگوریتم بهینه‌سازی کولونی مورچگان (ACO) می‌تواند پاسخ مناسبی برای جستجو در مسایل با فضای گراف از جمله انتخاب ویژگی باشد، در این کار یک الگوریتم تغییر یافته‌ی ACO در فاز فیلتر و به عنوان استراتژی جستجو معرفی شده است؛ اما اصلی‌ترین نوآوری این کار را می‌توان انتخاب مجموعه‌های مینیمم کاهش یافته‌ی ویژگی با اولین و دومین بهترین دقت دسته‌بند در نظر گرفت. ما برای تعیین کارآمدی روش ارایه شده، آن را بر روی ده مجموعه داده‌ی شناخته شده UCI آزمودیم. تحلیل نتایج به دست آمده حاکی از آن است که علی‌رغم روش‌های موجود، روش پیشنهادی ما قادر است هم­زمان دو شرط متضاد انتخاب ویژگی، یعنی افزایش دقت دسته بند و کاهش طول زیرمجموعه­های ویژگی را به دنبال داشته باشد.


1.      Jensen, R. and Shen, Q., "New approaches to fuzzy-rough feature selection", IEEE Transactions on Fuzzy Systems,  Vol. 17, No. 4, (2009), 824-838.

2.      Liu, H. and Yu, L., "Toward integrating feature selection algorithms for classification and clustering", IEEE Transactions on Knowledge and Data Engineering,  Vol. 17, No. 4, (2005), 491-502.

3.      Ferreira, A.J. and Figueiredo, M.A., "Efficient feature selection filters for high-dimensional data", Pattern Recognition Letters,  Vol. 33, No. 13, (2012), 1794-1804.

4.      Jensen, R. and Shen, Q., "Finding rough set reducts with ant colony optimization", in Proceedings of the UK workshop on computational intelligence. Vol. 1, (2003), 15-22.

5.      De Stefano, C., Fontanella, F., Marrocco, C. and Di Freca, A.S., "A ga-based feature selection approach with an application to handwritten character recognition", Pattern Recognition Letters,  Vol. 35, (2014), 130-141.

6.      Zhai, L.-Y., Khoo, L.-P. and Fok, S.-C., "Feature extraction using rough set theory and genetic algorithms—an application for the simplification of product quality evaluation", Computers & Industrial Engineering,  Vol. 43, No. 4, (2002), 661-676.

7.      Wang, X., Yang, J., Teng, X., Xia, W. and Jensen, R., "Feature selection based on rough sets and particle swarm optimization", Pattern Recognition Letters,  Vol. 28, No. 4, (2007), 459-471.

8.      Moaref, A. and Naeini, V.S., "A fuzzy-rough approach for finding various minimal data reductions using ant colony optimization", Journal of Intelligent & Fuzzy Systems,  Vol. 26, No. 5, (2014), 2505-2513.

9.      Mahdizadeh, M. and Eftekhari, M., "Proposing a novel cost sensitive imbalanced classification method based on hybrid of new fuzzy cost assigning approaches, fuzzy clustering and evolutionary algorithms", International Journal of Engineering-Transactions B: Applications,  Vol. 28, No. 8, (2015), 1160-1169.

10.    Shaeiri, Z. and Ghaderi, R., "Modification of the fast global k-means using a fuzzy relation with application in microarray data analysis", International Journal of Engineering-Transactions C: Aspects,  Vol. 25, No. 4, (2012), 283-291.

11.    Hamidi, H. and Daraei, A., "Analysis of pre-processing and post-processing methods and using data mining to diagnose heart diseases", International Journal of Engineering-Transactions A: Basics,  Vol. 29, No. 7, (2016), 921-930.

12.    Hamidzadeh, J., Zabihimayvan, M. and Sadeghi, R., "Detection of web site visitors based on fuzzy rough sets", Soft Computing,  (2017), 1-14.

13.    Dorigo, M. and Gambardella, L.M., "Ant colony system: A cooperative learning approach to the traveling salesman problem", IEEE Transactions on Evolutionary Computation,  Vol. 1, No. 1, (1997), 53-66.

14.    Costa, D. and Hertz, A., "Ants can colour graphs", Journal of the Operational Research Society,  Vol. 48, No. 3, (1997), 295-305.

15.    Merkle, D. and Middendorf*, M., "On solving permutation scheduling problems with ant colony optimization", International Journal of Systems Science,  Vol. 36, No. 5, (2005), 255-266.

16.    Okdem, S. and Karaboga, D., "Routing in wireless sensor networks using ant colony optimization", in Adaptive Hardware and Systems. AHS. First NASA/ESA Conference on, IEEE., (2006), 401-404.

17.    Maji, P. and Garai, P., "On fuzzy-rough attribute selection: Criteria of max-dependency, max-relevance, min-redundancy, and max-significance", Applied Soft Computing,  Vol. 13, No. 9, (2013), 3968-3980.

18.    Hu, Q., Yu, D., Xie, Z. and Liu, J., "Fuzzy probabilistic approximation spaces and their information measures", IEEE Transactions on Fuzzy Systems,  Vol. 14, No. 2, (2006), 191-201.

19.    Hu, Q., Yu, D. and Xie, Z., "Information-preserving hybrid data reduction based on fuzzy-rough techniques", Pattern Recognition Letters,  Vol. 27, No. 5, (2006), 414-423.

20.    Lee, T.T., "An infornation-theoretic analysis of relational databases—part i: Data dependencies and information metric", IEEE Transactions on Software Engineering,  Vol., No. 10, (1987), 1049-1061.

21.    Pawlak, Z., "Rough sets: Theoretical aspects of reasoning about data, Springer Science & Business Media,  Vol. 9,  (2012).

22.    Li, J., Mei, C. and Lv, Y., "A heuristic knowledge-reduction method for decision formal contexts", Computers & Mathematics with Applications,  Vol. 61, No. 4, (2011), 1096-1106.

23.             Dai, J. and Xu, Q., "Attribute selection based on information gain ratio in fuzzy rough set theory with application to tumor classification", Applied Soft Computing,  Vol. 13, No. 1, (2013), 211-221.

Download PDF 

International Journal of Engineering
E-mail: office@ije.ir
Web Site: http://www.ije.ir