Abstract
Aim: Purpose of the research is comparison of Variable Elimination Algorithm with an accurate feature elimination strategy for Ensemble Learning using NSL-KDD dataset.
Materials and Methods: Accuracy is analyzed for feature elimination. Classification of feature elimination is done using ensemble learning whose size of sample (N=34) as well as Variable Elimination Algorithm of (N=34) whose size of sample produced with G-power value 80%.
Results: Ensemble Learning accuracy becomes 82.74% and is more than Variable Elimination Algorithm with accuracy 76.62%. Significance value becomes 0.033 (p<0.05) indicating the performance of proposed work has significance.
Conclusion: Ensemble Learning performs whose accuracy is 82.74% when compared to Variable Elimination Algorithm of accuracy 76.62% along with the subset of variables in marginal distribution.