Fitting or training in machine learning is the process by which a model learns from input data. Fitting is another word for training an ML model. Besides the ideal best fit or good fit, a model can get overfitted when overfitting occurs or it can get underfitted when underfitting occurs.