{"id":13934956,"url":"https://github.com/arogozhnikov/infiniteboost","last_synced_at":"2025-06-12T23:35:47.318Z","repository":{"id":87372982,"uuid":"92594611","full_name":"arogozhnikov/infiniteboost","owner":"arogozhnikov","description":"InfiniteBoost: building infinite ensembles with gradient descent","archived":false,"fork":false,"pushed_at":"2018-09-17T21:11:59.000Z","size":4874,"stargazers_count":184,"open_issues_count":2,"forks_count":23,"subscribers_count":11,"default_branch":"master","last_synced_at":"2025-04-14T03:37:32.344Z","etag":null,"topics":["experiments","gradient-boosting","machine-learning","random-forest","research"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/arogozhnikov.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null}},"created_at":"2017-05-27T11:30:05.000Z","updated_at":"2024-09-14T16:11:34.000Z","dependencies_parsed_at":"2023-03-27T13:05:30.897Z","dependency_job_id":null,"html_url":"https://github.com/arogozhnikov/infiniteboost","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/arogozhnikov/infiniteboost","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arogozhnikov%2Finfiniteboost","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arogozhnikov%2Finfiniteboost/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arogozhnikov%2Finfiniteboost/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arogozhnikov%2Finfiniteboost/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/arogozhnikov","download_url":"https://codeload.github.com/arogozhnikov/infiniteboost/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arogozhnikov%2Finfiniteboost/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":259549940,"owners_count":22875177,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["experiments","gradient-boosting","machine-learning","random-forest","research"],"created_at":"2024-08-07T23:01:19.777Z","updated_at":"2025-06-12T23:35:47.276Z","avatar_url":"https://github.com/arogozhnikov.png","language":"Jupyter Notebook","funding_links":[],"categories":["Jupyter Notebook","2010"],"sub_categories":[],"readme":"\u003cimg src='https://github.com/arogozhnikov/infiniteboost/blob/master/infiniteboost.png' width=200 align=right /\u003e\n\n# InfiniteBoost\n\nCode for a paper  \u003cbr /\u003e\nInfiniteBoost: building infinite ensembles with gradient descent ([arXiv:1706.01109](https://arxiv.org/abs/1706.01109)). \u003cbr /\u003e\n[A. Rogozhnikov](https://github.com/arogozhnikov), [T. Likhomanenko](https://github.com/tlikhomanenko)\n\n## Description\n\n**InfiniteBoost** is an approach to building ensembles which combines best sides of random forest and gradient boosting. \n\nTrees in the ensemble encounter mistakes done by previous trees (as in gradient boosting), \nbut due to modified scheme of encountering contributions\nthe ensemble converges to the limit, thus avoiding overfitting (just as random forest).\n\n\u003cimg src='https://github.com/arogozhnikov/infiniteboost/blob/master/research/plots/rocauc_higgs.png' width=400 /\u003e\u003cimg src='https://github.com/arogozhnikov/infiniteboost/blob/master/research/plots/forest_longrun_real-sim.png' width=400 /\u003e \n\nLeft: InfiniteBoost with automated search of capacity vs gradient boosting with different learning rates (shrinkages), right: random forest vs InfiniteBoost with small capacities.\n\nMore plots of comparison in research notebooks and in research/plots directory.\n\n## Reproducing research\n\nResearch is performed in [jupyter notebooks](http://jupyter.org/)\n(if you're not familiar, read [why Jupyter notebooks are awesome](http://arogozhnikov.github.io/2016/09/10/jupyter-features.html)).\n\nYou can use the [docker image](https://docs.docker.com/engine/reference/commandline/images/) `arogozhnikov/pmle:0.01` from [docker hub](https://hub.docker.com/). \nDockerfile is stored in this repository (ubuntu 16 + basic [sklearn](https://github.com/scikit-learn/scikit-learn) stuff).\n\nTo run the environment (sudo is needed on Linux):\n```bash\nsudo docker run -it --rm -v /YourMountedDirectory:/notebooks -p 8890:8890 arogozhnikov/pmle:0.01\n```\n(and open [`localhost:8890`](http://localhost:8890/) in your browser).\n\n\n## InfiniteBoost package\n\nSelf-written minimalistic implementation of trees as used for experiments against boosting.\n\nSpecific implementation was used to compare with random forest and based on the trees from scikit-learn package. \n\nCode written in python 2 (expected to work with python 3, but not tested), some critical functions in fortran, so you need `gfortran + openmp` installed \nbefore installing the package (or simply use docker image).\n\n```bash\npip install numpy\npip install .\n# testing (optional)\ncd tests \u0026\u0026 nosetests .\n```\n\nYou can use implementation of trees from the package for your experiments, in this case please cite InfiniteBoost paper.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Farogozhnikov%2Finfiniteboost","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Farogozhnikov%2Finfiniteboost","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Farogozhnikov%2Finfiniteboost/lists"}