High R2_score on training data and very low on test data

I’m working on a dataset for load forecasting. I used linear regression for this problem .when I fit my model in my training data, it gave r2 score 1 which is quite high but when I tested this model for the test data I got 0.14 R2 score. what is the possible problem in my , model and what should be done to resolve this problem ?

Hi @_umesh, your model is overfitting on the training data. How many data points do you have for training and test sets?

Try using some regularization techniques and also check if the assumptions of the linear regression model are not violated. I believe the following article should help

© Copyright 2013-2019 Analytics Vidhya