The first method to help prevent overfitting is to train or fit your model on a portion of your dataset and then test or evaluate your model on a different portion of your dataset. Training your model generally consists of parameterizing one or more functions that make up your model, such that the functions that predict what you are trying to predict. Then, you can evaluate this trained model using one or more of the evaluation metrics that we discussed previously. The important thing here is that you do not want to test/evaluate your model on the same data that is used to train your model.
By reserving part of your data for testing, you are simulating the scenario in which your model sees new data. That is, the model ...
