Combination of Resampling Based Lasso Feature Selection and Ensembles of Regularized Regression Models
In high-dimensional data, the performance of various classifiers is largely dependent on the selection of important features. Most of the individual classifiers using existing feature selection (FS) methods do not perform well for highly correlated data. Obtaining important features using the FS method and selecting the best performing classier is a challenging task in high throughput data. In this research, we propose a combination of resampling based least absolute shrinkage and selection operator (LASSO) feature selection (RLFS) and ensembles of regularized regression models (ERRM) capable of handling data with the high correlation structures. The ERRM boosts the prediction accuracy with the top-ranked features obtained from RLFS. The RLFS utilizes the LASSO penalty with sure independence screening condition to select the top k ranked features. The ERRM includes ve individual penalty-based methods: LASSO, adaptive LASSO (ALASSO), elastic net (ENET), smoothly clipped absolute deviations (SCAD), and minimax concave penalty (MCP). It is built on the idea of bagging and rank aggregation. Upon performing simulation studies and applying to smokers cancer gene expression data, we demonstrated that the proposed combination of ERRM with RLFS achieved superior performance in accuracy and geometric mean.
Patil, Abhijeet R, "Combination of Resampling Based Lasso Feature Selection and Ensembles of Regularized Regression Models" (2019). ETD Collection for University of Texas, El Paso. AAI27671506.