Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Automating Machine Learning

Automating Machine Learning

Andreas Mueller

July 15, 2016
Tweet

More Decks by Andreas Mueller

Other Decks in Science

Transcript

  1. What? from automl import AutoClassifier clf = AutoClassifier().fit(X_train, y_train) >

    Current Accuracy: 70% (AUC .65) LinearSVC(C=1), 10sec > Current Accuracy: 76% (AUC .71) RandomForest(n_estimators=20) 30sec > Current Accuracy: 80% (AUC .74) RandomForest(n_estimators=500) 30sec
  2. Meta-Learning optimization Algorithm + Parameters Dataset 3 optimization Algorithm +

    Parameters Dataset 2 optimization Algorithm + Parameters Dataset 1
  3. Meta-Learning Meta-Features 1 optimization Algorithm + Parameters Dataset 3 optimization

    Algorithm + Parameters Dataset 2 optimization Algorithm + Parameters Dataset 1 Meta-Features 2 Meta-Features 3 ML model
  4. Meta-Learning Meta-Features 1 optimization Algorithm + Parameters Dataset 3 optimization

    Algorithm + Parameters Dataset 2 optimization Algorithm + Parameters Dataset 1 Meta-Features 2 Meta-Features 3 ML model New Dataset ML model Algorithm + Parameters
  5. Within Scikit-learn • GridSearchCV • RandomizedSearchCV • BayesianSearchCV (coming) •

    Searching over Pipelines (coming) • Built-in parameter ranges (coming)
  6. TODO Clean separation of: • Model Search Space • Pipeline

    Search Space • Optimization Method • Meta-Learning • Exploit prior knowledge better! • Usability • Runtime consideration
  7. TODO Clean separation of: • Model Search Space • Pipeline

    Search Space • Optimization Method • Meta-Learning • Exploit prior knowledge better! • Usability • Runtime consideration • Data subsampling
  8. 47 Material • Random Search for Hyper-Parameter Optimization (Bergstra, Bengio)

    • Efficient and Robust Automated Machine Learning (Feurer et al) [autosklearn] • http://automl.github.io/auto-sklearn/stable/ • Efficient Hyperparameter Optimization and Infinitely Many Armed Bandits (Lie et. al) [hyperband] https://arxiv.org/abs/1603.06560 • Scalable Bayesian Optimization Using Deep Neural Networks [Snoek et al]