Cross Validation

How to choose the machine learning model?

We can evalulate how the model performs with cross-validation

From the dataset, divide the training and validation dataset.

For example, use the first three blocks (75%) of data to train and the last block (25%) for test evalution. Keep track how well the method does with the test data.

Then, repeat for other combinations of train-test sets. Add or average the all results.

Also, 90% Train-10% Test can be used

Last updated

Was this helpful?