How do I increase my linear regression model’s ability to consider a feature into the results? I have two features that both contain the desired outputs from the prediction and I would like to somehow get them to overlap their results. When I check for corr they do not share the same correlating features and when I predict X on each their results are almost contrasting even though predicting ok feature 1 produces more accurate results.

What is something I can do when my linear regression model is not taking a feature that accounts for deviations into consideration for the prediction results?