Ensemble Learning Approach to Improve Existing Models

Agrim Mehra, Priyansha Tripathy, Ashhad Faridi, Ayes Chinmay1

1

Publication Date: 2019/12/17

Abstract: Machine learning is becoming an exciting field because it can be applied to the different problems we face in our daily lives. An essential function of machine learning is predicting the possibility of any future event. Improving this accuracy of prediction has always been one of the most challenging aspects of Machine Learning. In this paper we have tried to improve the accuracy by Combining the Ensemble Learning [1] approach of Bagging [2] and Boosting [3] with Linear regression. The problem of extrapolation in case of Bagging techniques is also explored and corrected using Ensemble methods. The combination will yield better results when used in the case of Bagging techniques as compared to the results given by individual models. Simulations of these algorithms are achieved in R and are further demonstrated in future sections.

Keywords: Ensemble Learning, Bagging, Boosting, Multiple Linear Regression, RMSE, Random Forest, Extrapolation.

DOI: No DOI Available

PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT19DEC198.pdf

REFERENCES

No References Available