Bayesian Tuning and Bandits
Project description
An open source project from Data to AI Lab at MIT.
A simple, extensible backend for developing auto-tuning systems.
Overview
Bayesian Tuning and Bandits is a simple, extensible backend for developing auto-tuning systems such as AutoML systems. It is currently being used in ATM (an AutoML system that allows tuning of classifiers) and MIT's system for the DARPA Data driven discovery of models program.
- Free software: MIT license
- Documentation: https://HDI-Project.github.io/BTB
- Homepage: https://github.com/HDI-Project/BTB
BTB is under active development. If you come across any issues, please report them here.
Installation
Install with pip
The easiest way to install BTB is using pip
.
pip install baytune
Build from source
You can also clone the repository and build it from source.
git clone git@github.com:HDI-Project/BTB.git
cd BTB
make install
Basic Usage
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 perform the following three steps in a loop.
-
Let the Tuner propose a new set of parameters
>>> parameters = tuner.propose() >>> parameters {'n_estimators': 297, 'max_depth': 3}
-
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
-
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 more detailed examples, check scripts from the examples
folder.
Selectors
The selectors are intended to be used in combination with 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.ensemble import RandomForestClassifier
>>> from sklearn.svm import SVC
>>> models = {
... 'RF': RandomForestClassifier,
... 'SVC': SVC
... }
>>> from btb.selection import UCB1
>>> selector = UCB1(['RF', 'SVC'])
>>> tuners = {
... 'RF': GP([
... ('n_estimators', HyperParameter(ParamTypes.INT, [10, 500])),
... ('max_depth', HyperParameter(ParamTypes.INT, [3, 20]))
... ]),
... 'SVC': GP([
... ('c', HyperParameter(ParamTypes.FLOAT_EXP, [0.01, 10.0])),
... ('gamma', HyperParameter(ParamTypes.FLOAT, [0.000000001, 0.0000001]))
... ])
... }
Then we perform the following steps in a loop.
-
Pass all the obtained scores to the selector and let it decide which model to test.
>>> next_choice = selector.select({'RF': tuners['RF'].y, 'SVC': tuners['SVC'].y}) >>> next_choice 'RF'
-
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)
-
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)
References
If you use BTB, please consider citing the following work:
-
Laura Gustafson. Bayesian Tuning and Bandits: An Extensible, Open Source Library for AutoML. Masters thesis, MIT EECS, June 2018. (pdf)
@MastersThesis{Laura:2018, title = "Bayesian Tuning and Bandits: An Extensible, Open Source Library for AutoML", author = "Laura Gustafson", month = "May", year = "2018", url = "https://dai.lids.mit.edu/wp-content/uploads/2018/05/Laura_MEng_Final.pdf", type = "M. Eng Thesis", address = "Cambridge, MA", school = "Massachusetts Institute of Technology", }
# History
## 0.2.5
### Bug Fixes
* Issue #115: HyperParameter subclass instantiation not working properly
## 0.2.4
### Internal Improvements
* Issue #62: Test for `None` in `HyperParameter.cast` instead of `HyperParameter.__init__`
### Bug fixes
* Issue #98: Categorical hyperparameters do not support `None` as input
* Issue #89: Fix the computation of `avg_rewards` in `BestKReward`
## 0.2.3
### Bug Fixes
* Issue #84: Error in GP tuning when only one parameter is present bug
* Issue #96: Fix pickling of HyperParameters
* Issue #98: Fix implementation of the GPEi tuner
## 0.2.2
### Internal Improvements
* Updated documentation
### Bug Fixes
* Issue #94: Fix unicode `param_type` caused error on python 2.
## 0.2.1
### Bug fixes
* Issue #74: `ParamTypes.STRING` tunables do not work
## 0.2.0
### New Features
* New Recommendation module
* New HyperParameter types
* Improved documentation and examples
* Fully tested Python 2.7, 3.4, 3.5 and 3.6 compatibility
* HyperParameter copy and deepcopy support
* Replace print statements with logging
### Internal Improvements
* Integrated with Travis-CI
* Exhaustive unit testing
* New implementation of HyperParameter
* Tuner builds a grid of real values instead of indices
* Resolve Issue #29: Make args explicit in `__init__` methods
* Resolve Issue #34: make all imports explicit
### Bug Fixes
* Fix error from mixing string/numerical hyperparameters
* Inverse transform for categorical hyperparameter returns single item
## 0.1.2
* Issue #47: Add missing requirements in v0.1.1 setup.py
* Issue #46: Error on v0.1.1: 'GP' object has no attribute 'X'
## 0.1.1
* First release.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for baytune-0.2.5-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 9a761e55b6ade3e3704fc64419b0df4c6fa976b12938969a2d87aecad23958e7 |
|
MD5 | cfd2eff498b57fce6f0a49a0a5f18dc2 |
|
BLAKE2b-256 | 47c83107ad844a656ffb703492e8687e9ce0f2bd0cec9ab5dae20c533c56cd3c |