Module advanced_models 

Advanced ML models: Random Forest, Gradient Boosting, and cross-validation.

Pure-Rust implementations for molecular property prediction beyond simple linear models. Includes:

• Random Forest (bagged decision trees)
• Gradient Boosted Trees (GBM / GBRT)
• K-fold cross-validation
• Model recalibration via isotonic regression

Structs

CrossValidationResult

Result of k-fold cross-validation.

GradientBoosting

Gradient Boosted Trees regressor.

GradientBoostingConfig

Configuration for Gradient Boosting.

IsotonicCalibrator

Isotonic regression for model recalibration. Fits a monotone non-decreasing function to prediction-target pairs.

RandomForest

Random Forest model: ensemble of bagged decision trees.

RandomForestConfig

Configuration for Random Forest.

TreeConfig

Configuration for tree building.

Enums

ModelType

Model type for cross-validation.

TreeNode

A single decision tree node (binary split).

Functions

build_tree

Build a decision tree from data.

cross_validate

Perform k-fold cross-validation.

train_gradient_boosting

Train a Gradient Boosting regressor (L2 loss / least squares).

train_random_forest

Train a Random Forest regressor.