Hybrid SVM - Random Forest Classi�cation System for Oral Cancer Screening using LIF Spectra

Singh, Rahul Kumar and Naik, Sarif Kumar and Gupta, Lalit and Balakrishnan, Srinivasan and Santhosh, C and Pai, Keerthilatha M (2008) Hybrid SVM - Random Forest Classi�cation System for Oral Cancer Screening using LIF Spectra. Conf Proc IEEE Eng Med Biol Soc. pp. 1-4.

[img] PDF
Hybrid_SVM_-_Random_Forest_Classifcation_System_for_Oral.pdf - Published Version
Restricted to Registered users only

Download (336kB) | Request a copy

Abstract

In this paper, a system for oral cancer screen- ing using Laser Induced Fluorescence(LIF) has been developed. A hybrid approach of classi�cation us- ing Support Vector Machine (SVM) and Random Forest (RF) classi�er's is proposed. Performance of the classi�er is evaluated using several features types such as Wavelet, DFT, LDFT, ILDFT, DCT, LDCT and Slopes features. The most discriminat- ing features are selected using Recursive Feature Elimination(RFE). Analysis of the problem of sub- set selection from SVM-RFE ranked list is also per- formed. The hybrid approach has been compared with stand-alone SVM, SVM-RFE and RF clas- si�ers. The proposed technique improves the per- formance of the classification system signi�cantly. The novelty of the approach lies in the way the most signi�cant features are exstracted in separate mod- ules to arrive at a decision and how the decision are then fused in an intelligent fashion to arrive at a final classifcation.

Item Type: Article
Subjects: Dentistry > MCODS Manipal > Oral Medicine and Radiology
Depositing User: KMC Manipal
Date Deposited: 18 Jul 2011 04:23
Last Modified: 05 Oct 2013 05:18
URI: http://eprints.manipal.edu/id/eprint/672

Actions (login required)

View Item View Item