Utils

This module contains the CrossValidator class, which facilitates the evaluation of QSAR models through cross-validation techniques. The class provides methods for creating cross-validation folds, calculating cross-validation scores, evaluating model performance across multiple metrics, and generating model predictions.

This module provides the Extractor class, designed for the efficient extraction and management of data from various CSV files for use in data analysis and modeling, particularly in QSAR studies. The Extractor simplifies the process of loading, accessing, and splitting datasets into features (X) and target (y) components, facilitating data handling and preprocessing steps in QSAR modeling workflows.

The HyperParameterOptimizer class is designed to streamline the process of hyperparameter optimization for QSAR models using the Optuna framework. This class is tailored to work seamlessly with models derived from the BaselineModel abstract class, leveraging Optuna’s efficient search capabilities to identify optimal hyperparameter settings based on specified evaluation criteria.

• Extractor
• Cross Validator
• Hyper Parameter Optimizer