Key Features

What you can do

Automatic Model Selection

Provides LazyClassifier and LazyRegressor classes to automatically build and compare multiple classification and regression models without parameter tuning.

Support for Numerical and Categorical Features

Handles both numerical and categorical data, including splitting categorical columns into low and high cardinality pipelines for better processing.

Integration with scikit-learn Pipelines

Easily integrates with scikit-learn pipelines to fit into existing machine learning workflows.

Model Performance Comparison

Generates tables comparing models on metrics such as accuracy, balanced accuracy, ROC AUC, and training time to facilitate model selection.

MLflow Integration

Supports MLflow for experiment tracking to manage and record model training runs.

Custom Evaluation Metrics and Predictions

Allows users to input custom evaluation metrics and returns predictions as DataFrames for further analysis.