Exploration of Multiple Linear Regression with Ensembling Schemes for Roof Fall Assessment using Machine Learning
M. Shyamala Devi1, Shakila Basheer2, Rincy Merlin Mathew3
1M. Shyamala Devi*, Associate Professor, Computer Science and Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Avadi, Chennai, Tamil Nadu, India.
2Shakila Basheer, Assistant Professor, Department of Information Systems, College of Computer and Information Sciences, Princess Nourah Bint Abdul Rahman university, Riyadh, Saudi Arabia.
3Rincy Merlin Mathew, Lecturer, Department of Computer Science, College of Science and Arts, Khamis Mushayt, King Khalid University, Abha, Asir, Saudi Arabia.
Manuscript received on September 14, 2019. | Revised Manuscript received on 26 September, 2019. | Manuscript published on October 10, 2019. | PP: 134-139 | Volume-8 Issue-12, October 2019. | Retrieval Number: L34741081219/2019©BEIESP | DOI: 10.35940/ijitee.L3474.1081219
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Roof fall of the building is the major threat to the society as it results in severe damages to the life of the people. Recently, engineers are focusing on the prediction of roof fall of the building in order to avoid the damage to the environment and people. Early prediction of Roof fall is the social responsibility of the engineers towards existence of health and wealth of the nation. This paper attempts to identify the essential attributes of the Roof fall dataset that are taken from the UCI Machine learning repository for predicting the existence of roof fall. In this paper, the important features are extorted from the various ensembling methods like Gradient Boosting Regressor, Random Forest Regressor, AdaBoost Regressor and Extra Trees Regressor. The extracted feature importance of each of the ensembling methods is then fitted with multiple linear regression to analyze the performance. The same extracted feature importance of each of the ensembling methods are subjected to feature scaling and then fitted with multiple linear regression to analyze the performance. The Performance analysis is done with the performance parameters such as Mean Squared Log Error (MSLE), Mean Absolute error (MAE), R2 Score, Mean Squared error (MSE) and Explained Variance Score (EVS). The execution is carried out using python code in Spyder Anaconda Navigator IP Console. Experimental results shows that before feature scaling, Extra Tree Regressor is found to be effective with the MSE of 0.06, MAE of 0.07, R2 Score of 87%, EVS of 0.89 and MSLE of 0.02 as compared to other ensembling methods. In the same way, after applying feature scaling, the feature importance extracted from the Extra Tree Regressor is found to be effective with the MSE of 0.04, MAE of 0.03, R2 Score of 96%, EVS of 0.9 and MSLE of 0.01 as compared to other ensembling methods.
Keywords: Machine Learning, Regression, MSLE, MAE, MSE and EVS.
Scope of the Article: Machine Learning