From the course: Automated ML.NET Training, Metrics, and Accuracy

Unlock the full course today

Join today to access over 24,700 courses taught by industry experts.

Cross-validation

Cross-validation - ML.NET Tutorial

From the course: Automated ML.NET Training, Metrics, and Accuracy

Start my 1-month free trial Buy for my team

Cross-validation

“

- [Instructor] Cross-validation is a technique for improving model accuracy. It's used for training and evaluation of the model. It's a technique for dividing data into partitions, also known as folds, and we use it to train multiple algorithms on these partitions. So it's effective for training models with smaller data sets. Afterwards, we can use the R-squared value to select the best model. To utilize cross-validation, we simply use the CrossValidate method for a given task. For example, regression or classification, they all have a CrossValidate method accessible within them. As a matter of fact, it's available in all ML tasks in ML.NET except for anomaly detection.

Contents