Mother tutorial

Here, you can find .ipynb files explaining how to effectively use Mother for chemical data procesing and model training:

Category	Tutorial Title	Description	Link
Data Processing	01_feature_selection.ipynb	Use a default feature selection pipeline in Mother.	View Tutorial
Data Processing	03_custom_preprocessing.ipynb	Design a custom preprocessing pipeline accordingly with Mother.	View Tutorial
Model Training	01_basic_regression.ipynb	Predict a continuous target value from molecular structure using CatboostRegressor implemented in Mother. This tutorial includes group k-fold cross-validation.	View Tutorial
Model Training	06_predict_predict_uncertainty_interface.ipynb	Demonstrates the unified predict and predict_uncertainty interface, including uncertainty outputs and model-specific behavior.	View Tutorial
Model Training	01_lasso_classification.ipynb	Classify molecules using a Lasso-based model in Mother.	View Tutorial
Optimisation	04_pipeline_with_settings.ipynb	Use MotherSettings to create a pipeline with Mother, including data preprocessing, feature generation and selection, and model training.	View Tutorial
Optimisation	01_custom_hyperparameter_optimization.ipynb	Customize hyperparameter optimization with Mother, following a baseline design of Optuna.	View Tutorial

Use MotherSettings to create a pipeline with Mother.

The pipeline can include all: data preprocessing, feature generation and selection, and model training.

Customise a hyperparameter optimisation with Mother.

This follows a baseline design of Optuna.