A Machine Learning Model to Prune Insignificant Attributes
Nidhi Agarwal, Neetika Arora Bajaj, Manjeet Kaur Ratan, Prakhar Deep
Abstract
In this research work, a machine learning model is proposed with only those features which are significantly contributing in prediction using multiple linear regression. The other insignificant features are pruned or eliminated using the concept of p-value. The research work will use p-value solely to determine which features are significant to the dataset and which are not so important. The results obtained are quite promising in terms of prediction power of the novel model as compared to the work done in literature till now.
Topics & Concepts
Computer scienceMachine learningValue (mathematics)Work (physics)Artificial intelligenceLinear regressionRegression analysisData miningEngineeringMechanical engineeringNeural Networks and ApplicationsStock Market Forecasting MethodsTime Series Analysis and Forecasting