optuna-hyperparameter-tuner

# Optuna Hyperparameter Tuner Optimize hyperparameters using Optuna with advanced search strategies, pruning, and visualization. ## Overview This skill provides comprehensive capabilities for hyperparameter optimization using Optuna, the state-of-the-art hyperparameter optimization framework. It supports various samplers, pruners, multi-objective optimization, and integration with popular ML frameworks. ## Capabilities ### Search Strategies - Tree-structured Parzen Estimator (TPE) - default, efficient - CMA-ES - for continuous parameters - Grid search - exhaustive - Random search - baseline - NSGAII - multi-objective optimization - QMC (Quasi-Monte Carlo) - low-discrepancy sampling ### Pruning Strategies - Median pruning - early stop underperformers - Hyperband (ASHA) - aggressive resource allocation - Percentile pruning - threshold-based - Successive Halving - efficient resource use - Wilcoxon pruning - statistical comparison ### Multi-Objective Optimization - Pareto front optimization - Multiple objective functions - Constraint handling - Trade-off visualization ### Study Management - Study persistence (SQLite, PostgreSQL, MySQL) - Study resumption - Parallel/distributed optimization - Trial importance analysis - Parameter relationship analysis ### Visualization - Optimization history - Parameter importance - Parallel coordinate plots - Slice plots - Contour plots ## Prerequisites ### Installation ```bash pip install optuna>=3.0.0 ``` ### Optional Dependencies `...