MATIH Platform is in active MVP development. Documentation reflects current implementation status.
13. ML Service & MLOps
Hyperparameter Tuning

Hyperparameter Tuning

The MATIH ML Service provides automated hyperparameter optimization using Ray Tune, supporting Bayesian optimization, grid search, random search, and population-based training strategies.


Tuning API

POST /api/v1/tuning/jobs
Content-Type: application/json
 
{
  "model_class": "sklearn.ensemble.GradientBoostingClassifier",
  "search_space": {
    "learning_rate": {"type": "loguniform", "min": 0.001, "max": 0.1},
    "n_estimators": {"type": "choice", "values": [100, 200, 500]},
    "max_depth": {"type": "randint", "min": 3, "max": 12}
  },
  "search_algorithm": "bayesian",
  "num_trials": 50,
  "metric": "val_accuracy",
  "mode": "max",
  "max_concurrent_trials": 4
}

Search Algorithms

AlgorithmDescriptionBest For
bayesianBayesian optimization (Optuna/HyperOpt)Small-medium search spaces
gridExhaustive grid searchDiscrete, small spaces
randomRandom samplingLarge search spaces
pbtPopulation-based trainingNeural network schedules

Source Files

FilePath
HyperparameterTunerdata-plane/ml-service/src/training/hyperparameter_tuner.py
Tuning APIdata-plane/ml-service/src/api/tuning.py