HYBRID ENSEMBLE GENE SELECTION ALGORITHM FOR IDENTIFYING BIOMARKERS FROM BREAST CANCER GENE EXPRESSION PROFILES?
Journal: International Journal of Computer Science and Mobile Computing - IJCSMC (Vol.2, No. 9)Publication Date: 2013-09-30
Authors : S. ARUNA L.V. NANDAKISHORE;
Page : 153-159
Keywords : Ada boost; Correlated feature selection; Filters; Gene expression profiles; Sequential floating search algorithms; Support vector machines; Wrappers;
Abstract
Breast cancer is one of the major health hazard in the world. DNA gene expression profiles plays an important role in identifying the biomarkers for cancer which not only help in accurate diagnosis of the disease, also in discovering drugs, minimizing the toxicity thus help in the effective management of the disease. In this paper we propose an algorithm for determining the biomarkers. Our hybrid ensemble gene selection algorithm was experimented over breast cancer gene expression data of 24481 genes. The algorithm selected a marker gene subset of eight genes with an accuracy and BER of 96.9% and 0.033 respectively.
Other Latest Articles
- An Improved Compiler Directed Power Optimization for Disk Based Systems using Back Propagation Neural Networks
- Secure Medical Data Transmission in Body Area Sensor Networks Using Dynamic Biometrics and Steganography
- An Application of the Fuzzy Cognitive Mappings to Study about the Problems Faced by the Sports Students in Colleges
- Improved BSP Clustering Algorithm for Social Network Analysis
- Optimization of Milling Operation Using Genetic and PSO Algorithm
Last modified: 2013-09-26 22:41:27