The Model garden module is based on the library LazyPredict, which allows users to select the best model for the task at hand, running it on multiple types of machine learning models and returning the performance metrics for each model tested on the input dataset.
The objective of the model garden library is also to determine, given a learning task, the best data preparation (for example undersampling or oversampling) strategy and the most relevant machine learning evaluation metrics for the specific use case.
The Model Garden is employed in the ML Utility Metrics module.
ClassificationSandbox()ClassificationSandbox.fit()RegressionSandbox()RegressionSandbox.fit()