Journal of the Institute of Electronics and Information Engineers (전자공학회논문지)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지
DOI QR코드

DOI QR Code

Binary Tree Architecture Design for Support Vector Machine Using Dynamic Time Warping

DTW를 이용한 SVM 기반 이진트리 구조 설계

  • Kang, Youn Joung (Department of Ocean System Engineering, Jeju National University) ;
  • Lee, Jaeil (Department of Ocean System Engineering, Jeju National University) ;
  • Bae, Jinho (Sonar Systems PMO, Agency for Defense Development) ;
  • Lee, Seung Woo (Sonar Systems PMO, Agency for Defense Development) ;
  • Lee, Chong Hyun (Department of Ocean System Engineering, Jeju National University)
  • 강윤정 (제주대학교 해양시스템공학과) ;
  • 이재일 (제주대학교 해양시스템공학과) ;
  • 배진호 (국방과학연구소 소나체계개발단) ;
  • 이승우 (국방과학연구소 소나체계개발단) ;
  • 이종현 (제주대학교 해양시스템공학과)
  • Received : 2013.11.11
  • Accepted : 2014.05.25
  • Published : 2014.06.25
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, we propose the classifier structure design algorithm using DTW. Proposed algorithm uses DTW result to design the binary tree architecture based on the SVM which classify the multi-class data. Design the binary tree architecture for Support Vector Machine(SVM-BTA) using the threshold criterion calculated by the sum columns in square matrix which components are the reference data from each class. For comparison the performance of the proposed algorithm, compare the results of classifiers which binary tree structure are designed based on database and k-means algorithm. The data used for classification is 333 signals from 18 classes of underwater transient noise. The proposed classifier has been improved classification performance compared with classifier designed by database system, and probability of detection for non-biological transient signal has improved compare with classifiers using k-means algorithm. The proposed SVM-BTA classified 68.77% of biological sound(BO), 92.86% chain(CHAN) the mechanical sound, and 100% of the 6 kinds of the other classes.

본 논문은 DTW 결과를 이용하여 분류기 구조를 설계하는 알고리즘을 제안한다. 제안된 알고리즘은 다수 클래스의 데이터를 분류하기 위한 SVM 기반 이진트리 구조를 설계하는데 있어 DTW 결과를 이용한다. 각 클래스에 대한 데이터를 DTW의 입력으로 하여 얻어진 결과행렬의 열의 합을 이용하여 계산된 임계치를 기준으로 SVM 기반 이진트리 구조(SVM-BTA)를 설계한다. 제안된 알고리즘의 성능 비교를 위해 데이터베이스와 k-means 알고리즘을 이용한 이진트리 구조의 분류 결과를 비교한다. 분류에 사용된 데이터는 수중과도소음 데이터베이스의 18개 클래스 333개의 데이터이다. 제안된 분류기는 데이터베이스의 체계를 이용한 분류기에 비해 분류성능이 향상되었고, k-means 알고리즘을 이용한 분류기에 비해 비 생물소음의 검출 확률이 향상되었다. 제안된 SVM-BTA는 생물 소음(BO) 68.77%, 기계 소음인 체인(CHAN) 92.86%, 그 외의 기계 소음 및 음향학적 소음, 기타소음의 6종은 100%로 분류한다.

Keywords

References

  1. C. W. Clark, R. Charif, S. Mitchell and J. Colby, "Distribution and behavior of the bowhead whale, Balaena mysticetus, based on analysis of acoustic data collected during the 1993 spring migration off Point Barrow, Alaska," Report-International Whaling Commission 46, pp. 541-554, Dublin, Ireland, June 1995.
  2. S. P. Van IJsselmuide and S. P. Beerens. "Detection and classification of marine mammals using an LFAS system," Canadian Acoustics, Vol. 32, no. 2, pp. 93-106, June 2004.
  3. Q. Q. Huynh, L. N. Cooper, N. Intrator and H. Shouval, "Classification of underwater mammals using feature extraction based on time-frequency analysis and BCM theory," IEEE Trans. Signal Processing, Vol. 46, no. 5, pp. 1202-1207, May 1998. https://doi.org/10.1109/78.668783
  4. Y. J. Kang, C. H. Lee, J. Bae, S. Lee, and S. W. Lee, "SVM classifier design using DTW preprocessing," in Proc. of IEEK summer Conf., pp. 556-559, Jeju, Korea, July 2013.
  5. F. Takahashi and S. Abe, "Decision-tree-based multiclass support vector machines," in Proc. of IEEE Conf. on Neural Information Processing, Vol. 3. pp. 1418-1422, Orchid Country Club, Singapore, November 2002.
  6. T. G. Lim, I. H. Kim, T. H. Kim, and K. S. Bae, "Frame Based Classification of Underwater Transient Signal Using MFCC Feature Vector and Neural Network," in Proc. of IEEK summer Conf., pp. 883-884, Jeju, Korea, June 2008.
  7. K. S. R. Murty and B. Yegnanarayana. "Combining evidence from residual phase and MFCC features for speaker recognition," IEEE Signal Processing Letters, Vol. 13, no. 1, pp. 52-55, January 2006. https://doi.org/10.1109/LSP.2005.860538
  8. M. Martinez-Ramon and C. Christodouloui, Support vector machines for antenna array processing and electromagnetics, Morgan & Claypool, pp. 7-32, 2006.
  9. S. Cheong, S. H. Oh, and S. Lee, "Support vector machines with binary tree architecture for multi-class classification." Neural Information Processing Letters and Reviews, Vol. 2, no. 3, pp. 47-51, March 2004.
  10. H. Han, Pattern recognition, Hanbit Media, pp. 419-424, 2009.
  11. J. R. Dellar, J. G. Proakis, and J. H. L. Hansem, Discrete-Time Processing of Speech Signals, Macmillan Publishing Company, pp.1380-1397, 1993.