| Name | valda JSON |
| Version |
0.1.10
JSON |
| download |
| home_page | |
| Summary | A Data Valuation Package for Machine Learning |
| upload_time | 2023-01-28 15:45:47 |
| maintainer | |
| docs_url | None |
| author | |
| requires_python | >=3.6 |
| license | |
| keywords |
data
valuation
|
| VCS |
|
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
# Valda
## Introduction
Valda is a Python package for data valuation in machine learning. If you are interested in
- analyzing the contribution of individual training examples to the final classification performance, or
- identifying some noisy examples in the training set,
you may be interested in the functions provided by this package.
The current version supports five different data valuation methods. It supports all the classifiers from Sklearn for valuation, and also user-defined classifier using PyTorch.
- Leave-one-out (LOO),
- Data Shapley with the TMC algorithm (TMC-Shapley) from [Ghorbani and Zou (2019)](https://proceedings.mlr.press/v97/ghorbani19c.html),
- Beta Shapley from [Kwon and Zou (2022)](https://arxiv.org/abs/2110.14049)
- Class-wise Shapley (CS-Shapley) from [Schoch et al. (2022)](https://arxiv.org/abs/2211.06800)
- Influence Function (IF) from [Koh and Liang (2017)](https://arxiv.org/abs/1703.04730)
- IF only works with the classifiers built with PyTorch, because it requires gradient computation.
- v0.1.8 only support the first-order gradient computation, and we will add the second-order computation soon.
## Tutorial
Please checkout a simple tutorial on [Google Colab](https://colab.research.google.com/drive/1agsMNqZan-3RnJLQtBGATRHHWYMe7C9H?usp=sharing), for how to use this package.
Raw data
{
"_id": null,
"home_page": "",
"name": "valda",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.6",
"maintainer_email": "",
"keywords": "data valuation",
"author": "",
"author_email": "Yangfeng Ji <yangfeng@virginia.edu>, Stephanie Schoch <sns2gr@virginia.edu>",
"download_url": "https://files.pythonhosted.org/packages/21/78/7bdf61979f801314e997771b2def6ecf90bffcab9ce8933a091dd47a04d4/valda-0.1.10.tar.gz",
"platform": null,
"description": "# Valda\n\n## Introduction\n\nValda is a Python package for data valuation in machine learning. If you are interested in \n\n- analyzing the contribution of individual training examples to the final classification performance, or \n- identifying some noisy examples in the training set, \n\nyou may be interested in the functions provided by this package.\n\n\nThe current version supports five different data valuation methods. It supports all the classifiers from Sklearn for valuation, and also user-defined classifier using PyTorch. \n\n- Leave-one-out (LOO), \n- Data Shapley with the TMC algorithm (TMC-Shapley) from [Ghorbani and Zou (2019)](https://proceedings.mlr.press/v97/ghorbani19c.html), \n- Beta Shapley from [Kwon and Zou (2022)](https://arxiv.org/abs/2110.14049)\n- Class-wise Shapley (CS-Shapley) from [Schoch et al. (2022)](https://arxiv.org/abs/2211.06800)\n- Influence Function (IF) from [Koh and Liang (2017)](https://arxiv.org/abs/1703.04730)\n - IF only works with the classifiers built with PyTorch, because it requires gradient computation.\n - v0.1.8 only support the first-order gradient computation, and we will add the second-order computation soon.\n\n\n\n\n## Tutorial\n\nPlease checkout a simple tutorial on [Google Colab](https://colab.research.google.com/drive/1agsMNqZan-3RnJLQtBGATRHHWYMe7C9H?usp=sharing), for how to use this package. \n",
"bugtrack_url": null,
"license": "",
"summary": "A Data Valuation Package for Machine Learning",
"version": "0.1.10",
"split_keywords": [
"data",
"valuation"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "0c3e3ebbb4f310898ca87bd9e29a31e85155b838a017a5738e59666efb5deff6",
"md5": "e206c4630105d34f40d1687516f35181",
"sha256": "aa8cdddccd296cabe097aee7cee45dbc23b97f52622da50b6167b5ecb9845792"
},
"downloads": -1,
"filename": "valda-0.1.10-py3-none-any.whl",
"has_sig": false,
"md5_digest": "e206c4630105d34f40d1687516f35181",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.6",
"size": 14253,
"upload_time": "2023-01-28T15:45:46",
"upload_time_iso_8601": "2023-01-28T15:45:46.359639Z",
"url": "https://files.pythonhosted.org/packages/0c/3e/3ebbb4f310898ca87bd9e29a31e85155b838a017a5738e59666efb5deff6/valda-0.1.10-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "21787bdf61979f801314e997771b2def6ecf90bffcab9ce8933a091dd47a04d4",
"md5": "5b96727a828357185716a0423eaebddd",
"sha256": "67e42c8446ea2e0baa42a1da769bf9d2758705205107d772f140ef4731a8e585"
},
"downloads": -1,
"filename": "valda-0.1.10.tar.gz",
"has_sig": false,
"md5_digest": "5b96727a828357185716a0423eaebddd",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.6",
"size": 11431,
"upload_time": "2023-01-28T15:45:47",
"upload_time_iso_8601": "2023-01-28T15:45:47.911165Z",
"url": "https://files.pythonhosted.org/packages/21/78/7bdf61979f801314e997771b2def6ecf90bffcab9ce8933a091dd47a04d4/valda-0.1.10.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-01-28 15:45:47",
"github": false,
"gitlab": false,
"bitbucket": false,
"lcname": "valda"
}
