Differential Evolution (DE) based Multiple Regression Model for Classification
B. Thanga Parvathi, S. MercyShalinie · 2014
This paper proposes the building of a classifier model based on Multiple Regression Linear Models (MRLM). The performance of the regression models are estimated using Least Square Estimation Technique (LSE) and can be improved by optimizing its coefficients. The present work proposes a unique approach of using Differential Evolution (DE) for optimizing the coefficients of developed MRLM. Few datasets with different attributes are collected from UCI Machine Learning repository used for developing mathematical models for MRLM and the accuracy of the models are estimated using LSE and the coefficients are optimized using DE. The interrelationship of attributes of the model with respect to the class membership is used for prediction. The results are compared for validating the improved performance and found that the optimized coefficient of the developed mathematical model outperforms the traditional regression model in terms of classification accuracy. /* Style Definitions */ table.MsoNormalTable {mso-style-name:Table Normal; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:; mso-padding-alt:0cm 5.4pt 0cm 5.4pt; mso-para-margin:0cm; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:Times New Roman; mso-fareast-font-family:Times New Roman; mso-ansi-language:#0400; mso-fareast-language:#0400; mso-bidi-language:#0400;}