I have built the model using train data with good R squared.
In sample submissions, I saw the Price variable. I combined the test data and sample submission (it has only one variable named Price). I considered the Price variable in sample submissions as a target variable for test data. Is it correct to take that as an actual target for test data?
Then I predicted and found the prediction accuracy using the actual target in test (taken from sample submissions) and predicted values. The accuracy is low. It seems overfitting issue. May I know, What is the remedy?