Hi Madhu Sudhan,

Yes, I agree that using pd.get_dummies() before train test split will solve column mismatch issue. But this may cause issue for future unseen points, as the characteristics of encoding are not saved.

Thanks