Decision Trees Ensembles
Project description
Woods: Decision Tree Ensembles
Currently implemented algorithms:
- Partially randomized decision tree (variance minimization).
- Gradient Boosting of decision trees (MSE minimization).
- Average ensemble of GBM.
- Deep Gradient Boosting (of Average ensembles of GBM).
TODO
- Implement median-split, best-split decision tree;
- Provide optional min&max search based on pre-sorting (find min&max of
array[indices]
); - Add different loss-functions, ranking support.
Installation
Build environment
- Install
rustup
. - Set up nightly toolchain:
rustup toolchain install nightly
rustup default nightly
Install Python extension
Run setup.py
:
python setup.py install --user
Note that --user
option is used to install package locally.
Build documentation
Go to rust
dir and run:
cargo doc --lib
Docs will be placed in target/doc/woods/index.html
.
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
woods-0.1.0.tar.gz
(1.7 kB
view hashes)
Built Distributions
woods-0.1.0-cp37-cp37m-win_amd64.whl
(299.5 kB
view hashes)
woods-0.1.0-cp36-cp36m-win32.whl
(364.4 kB
view hashes)
Close
Hashes for woods-0.1.0-py3.6-macosx-10.9-x86_64.egg
Algorithm | Hash digest | |
---|---|---|
SHA256 | ccaa80b0cbc795f0ccc1d2e711f7f697da94e00f27090f84e74e0de3a1dcdaf9 |
|
MD5 | ccb9eb391b95b4b78610c7c596b0ac28 |
|
BLAKE2b-256 | 60a2922fcf53e8ef75b5148a88d0ea6b4e3fc0f5e01a5b4808b562ffd7ff468d |
Close
Hashes for woods-0.1.0-cp37-cp37m-win_amd64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 7ab8d251bc2f4c2ddcd60588fed6b4c7f16b277b4f7e3fe85ef44e19e695545f |
|
MD5 | 2b1ff0e73804e862ebe8018b03a7cc57 |
|
BLAKE2b-256 | 05172670f6784dcb592bf986d4d6147e453b78be371fb97fc84f3f46bd6103b4 |
Close
Hashes for woods-0.1.0-cp36-cp36m-win32.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6f247c9fc175d01caaf38be0a9d497d344760dad31a375023438de87a0ed40b2 |
|
MD5 | 7b408121797f478d145293fe4c059bc1 |
|
BLAKE2b-256 | 17d5d8bbc70b72d9bb0705bc689436635386a13e3de1c7311f8141d55b76bf7f |
Close
Hashes for woods-0.1.0-cp36-cp36m-macosx_10_15_x86_64.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 98a89903a3c229249165842ae9727f96ae8b2fb1985bba743bb11e88d0a8329b |
|
MD5 | b0fad7694b3c408930b3457998b20e63 |
|
BLAKE2b-256 | a1940410cda043778bbd2eb610c1931662d562c957fae0b137711859d41bf24b |