BTB: Bayesian Tuning and Bandits

Smart selection of hyperparameters

• Free software: MIT license
• Documentation: https://HDI-Project.github.io/BTB

Overview

Bayesian Tuning and Bandits is a simple, extensible Auto Machine Learning system that automates model selection and hyperparameter tuning.

Submodules

• selection defines Selectors: classes for choosing from a set of discrete options with multi-armed bandits
• tuning defines Tuners: classes with a fit/predict/propose interface for suggesting sets of hyperparameters

Tuners

Tuners are specifically designed to speed up the process of selecting the optimal hyper parameter values for a specific machine learning algorithm.

This is done by following a Bayesian Optimization approach and iteratively:

• letting the tuner propose new sets of hyper parameter
• fitting and scoring the model with the proposed hyper parameters
• passing the score obtained back to the tuner

At each iteration the tuner will use the information already obtained to propose the set of hyper parameters that it considers that have the highest probability to obtain the best results.

Selectors

Selectors apply multiple strategies to decide which models or families of models to train and test next based on how well thay have been performing in the previous test runs. This is an application of what is called the Multi-armed Bandit Problem.

The process works by letting know the selector which models have been already tested and which scores they have obtained, and letting it decide which model to test next.

Installation

Install with pip

BTB is not pulished in PyPi yet, but you can already install the latest release using pip

pip install -e git+https://github.com/HDI-Project/[email protected]#egg=btb

Install from sources

You can also clone the repository and install it from sources

git clone [email protected]:HDI-Project/BTB.git
cd BTB
make install

Usage examples

Tuners

In order to use a Tuner we will create a Tuner instance indicating which parameters we want to tune, their types and the range of values that we want to try

>>> from btb.tuning import GP
>>> from btb import HyperParameter, ParamTypes
>>> tunables = [
... ('n_estimators', HyperParameter(ParamTypes.INT, [10, 500])),
... ('max_depth', HyperParameter(ParamTypes.INT, [3, 20]))
... ]
>>> tuner = GP(tunables)

Then we into a loop and perform three steps:

1. Let the Tuner propose a new set of parameters

>>> parameters = tuner.propose()
>>> parameters
{'n_estimators': 297, 'max_depth': 3}

2. Fit and score a new model using these parameters

>>> model = RandomForestClassifier(**parameters)
>>> model.fit(X_train, y_train)
RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',
            max_depth=3, max_features='auto', max_leaf_nodes=None,
            min_impurity_decrease=0.0, min_impurity_split=None,
            min_samples_leaf=1, min_samples_split=2,
            min_weight_fraction_leaf=0.0, n_estimators=297, n_jobs=1,
            oob_score=False, random_state=None, verbose=0,
            warm_start=False)
>>> score = model.score(X_test, y_test)
>>> score
0.77

3. Pass the used parameters and the score obtained back to the tuner

tuner.add(parameters, score)

At each iteration, the Tuner will use the information about the previous tests to evaluate and propose the set of parameter values that have the highest probability of obtaining the highest score.

For a more detailed example, check scripts from the examples folder.

Selectors

The selectors are intended to be used in combination with the Tuners in order to find out and decide which model seems to get the best results once it is properly fine tuned.

In order to use the selector we will create a Tuner instance for each model that we want to try out, as well as the selector instance.

>>> from sklearn.svm import SVC
>>> models = {
...     'RF': RandomForestClassifier,
...     'SVC': SVC
... }
>>> from btb.selection import UCB1
>>> selector = UCB1(['RF', 'SVM'])
>>> tuners = {
...     'RF': GP([
...         ('n_estimators', HyperParameter(ParamTypes.INT, [10, 500])),
...         ('max_depth', HyperParameter(ParamTypes.INT, [3, 20]))
...     ]),
...     'SVM': GP([
...         ('c', HyperParameter(ParamTypes.FLOAT_EXP, [0.01, 10.0])),
...         ('gamma', HyperParameter(ParamTypes.FLOAT, [0.000000001, 0.0000001]))
...     ])
... }

Then, we will go into a loop and, at each iteration, perform the steps:

1. Pass all the obtained scores to the selector and let it decide which model to test

>>> next_choice = selector.select({'RF': tuners['RF'].y, 'SVM': tuners['SVM'].y})
>>> next_choice
'RF'

2. Obtain a new set of parameters from the indicated tuner and create a model instance

>>> parameters = tuners[next_choice].propose()
>>> parameters
{'n_estimators': 289, 'max_depth': 18}
>>> model = models[next_choice](**parameters)

3. Evaluate the score of the new model instance and pass it back to the tuner

>>> model.fit(X_train, y_train)
RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',
            max_depth=18, max_features='auto', max_leaf_nodes=None,
            min_impurity_decrease=0.0, min_impurity_split=None,
            min_samples_leaf=1, min_samples_split=2,
            min_weight_fraction_leaf=0.0, n_estimators=289, n_jobs=1,
            oob_score=False, random_state=None, verbose=0,
            warm_start=False)
>>> score = model.score(X_test, y_test)
>>> score
0.89
>>> tuners[next_choice].add(parameters, score)