Struct MLTrainer 

pub struct MLTrainer { /* private fields */ }

ML model trainer

Implementations

impl MLTrainer

pub fn new( n_estimators: usize, max_depth: Option<usize>, max_features: usize, ) -> Self

Create a new ML trainer

Arguments

• n_estimators - Number of trees in Random Forest
• max_depth - Maximum depth of each tree (None for unlimited)
• max_features - Maximum TF-IDF features

Examples

use organizational_intelligence_plugin::ml_trainer::MLTrainer;

let trainer = MLTrainer::new(100, Some(20), 1500);

pub fn load_dataset<P: AsRef<Path>>(path: P) -> Result<TrainingDataset>

Load training dataset from JSON file

Arguments

• path - Path to training data JSON file

Returns

• Ok(TrainingDataset) if successful
• Err if file not found or invalid format

pub fn train(&self, dataset: &TrainingDataset) -> Result<TrainedModel>

Train ML classifier on training dataset

Arguments

• dataset - Training dataset with splits

Returns

• Ok(TrainedModel) with trained classifier
• Err if training fails

Examples

use organizational_intelligence_plugin::ml_trainer::MLTrainer;
use std::path::PathBuf;

let trainer = MLTrainer::new(100, Some(20), 1500);
let dataset = MLTrainer::load_dataset("training-data.json")?;
let model = trainer.train(&dataset)?;

pub fn evaluate( model: &TrainedModel, test_examples: &[TrainingExample], ) -> Result<f32>

Evaluate model on test set

Arguments

• model - Trained model
• test_examples - Test examples

Returns

• Test accuracy (0.0-1.0)

pub fn save_model<P: AsRef<Path>>(model: &TrainedModel, path: P) -> Result<()>

Save trained model to disk

Arguments

• model - Trained model
• path - Path to save model JSON

Returns

• Ok(()) if successful
• Err if save fails

pub fn load_model<P: AsRef<Path>>(path: P) -> Result<TrainedModel>

Load trained model from disk

Arguments

• path - Path to model JSON file

Returns

• Ok(TrainedModel) if successful
• Err if load fails

Trait Implementations

impl Default for MLTrainer

fn default() -> Self

Returns the “default value” for a type.