Articles on: Training models

Train/Test split

Train-Test Split is a technique in machine learning to divide a dataset into two parts: training data and testing data. The purpose is to use the training data to build a model, and then test the model's accuracy on the testing data.

For example, let's say you have a dataset of 1000 student exam scores. You could split it into a training set of 800 scores and a testing set of 200 scores. You would use the 800 scores to train a model, and then use the 200 scores to test the accuracy of the model.

A commonly used ratio for the train-test split is 80:20, meaning 80% of the data is used for training and 20% for testing. However, the ideal ratio depends on the size and quality of the dataset, and the type of model being used. A ratio of 60:40 or 70:30 can also work in some cases.

NextBrain allows you to set your training and test split. However, note that, the least train proportion you can have is 40% since model would bear a high risk of not learning the data below that number.

Updated on: 31/01/2023

Was this article helpful?

Share your feedback


Thank you!