mlfastopt 0.0.9b1

ML Fast Opt - Advanced ensemble optimization system for LightGBM hyperparameter tuning

MLFastOpt

MLFastOpt is a high-speed ensemble optimization system for Bayesian hyperparameter tuning of LightGBM models.

Features

• 🚀 Fast Optimization: Advanced Bayesian optimization algorithms.
• 🎯 LightGBM Focused: Automated ensemble creation and tuning.
• ⚙️ Simple Config: JSON-based configuration and Python-based search spaces.
• 📊 Rich Analytics: Built-in web dashboards and visualization tools.

Installation

pip install mlfastopt

Quick Start

Prerequisite: Input data must be preprocessed and numerical. Handle all categorical encoding (e.g., one-hot, label encoding) before using MLFastOpt.

1. Setup

Create the required directory structure:

mkdir -p config/hyperparameters data

2. Define Parameter Space

Create config/hyperparameters/my_space.py:

PARAMETERS = [
    {"name": "num_leaves", "type": "range", "bounds": [20, 200], "value_type": "int"},
    {"name": "learning_rate", "type": "range", "bounds": [0.01, 0.3], "value_type": "float", "log_scale": True},
    {"name": "n_estimators", "type": "range", "bounds": [100, 300], "value_type": "int"},
    # Add other LightGBM parameters as needed
]

def get_parameter_space():
    return PARAMETERS

3. Configure

Create my_config.json:

{
  "DATA_PATH": "data/your_dataset.parquet",
  "HYPERPARAMETER_PATH": "config/hyperparameters/my_space.py",
  "LABEL_COLUMN": "target",
  "FEATURES": ["feature1", "feature2"],
  "N_ENSEMBLE_GROUP_NUMBER": 5,
  "AE_NUM_TRIALS": 20,
  "PARALLEL_TRAINING": true,
  "N_JOBS": -1
}

4. Run

Execute optimization (ensure single-threading for LightGBM to avoid deadlocks):

export OMP_NUM_THREADS=1
python -m mlfastopt.cli --config my_config.json

Configuration Reference

Parameter	Description	Default
DATA_PATH	Path to dataset (CSV/Parquet).	Required
HYPERPARAMETER_PATH	Path to parameter space file.	Required
LABEL_COLUMN	Name of target column.	Required
FEATURES	List of feature names.	Required
N_ENSEMBLE_GROUP_NUMBER	Models per ensemble.	1
AE_NUM_TRIALS	Total optimization trials.	20
NUM_SOBOL_TRIALS	Initial random trials.	5
OPTIMIZATION_METRICS	Metric to maximize (soft_recall, soft_f1_score, etc).	soft_recall
SAVE_THRESHOLD_ENABLED	Save only models exceeding metric threshold.	false
ENABLE_DATA_IMPUTATION	Simple median/mode imputation.	false

Outputs

Results are saved to outputs/:

• runs/: Detailed logs and models for each run.
• best_trials/: JSON configurations of the best performing trials.
• visualizations/: Generated plots.