文件名称:feature-selection-for-machine-learning:在线课程机器学习功能选择的代码存储库
文件大小:1.47MB
文件格式:ZIP
更新时间:2024-05-27 13:56:24
JupyterNotebook
机器学习的特征选择-代码存储库 2018年2月发布 链接 目录 基本选择方法 删除常量功能 删除准恒定特征 删除重复的功能 相关特征选择 删除相关功能 基本选择方法+相关-管道 过滤方法:单变量统计方法 相互信息 卡方分布 阿诺娃 基本选择方法+统计方法-管道 筛选器方法:其他方法和指标 单变量roc-auc,mse等 KDD竞赛中使用的方法-2009 包装方法 前进功能选择 后退功能选择 详尽的特征选择 嵌入式方法:线性模型系数 逻辑回归系数 线性回归系数 正则化对系数的影响 基本选择方法+相关性+嵌入式-管道 嵌入式方法:套索 套索 基本选择方法+关联+套索-管道 嵌入式方法:树的重要性 随机森林衍生的特征重要性 树重要性+递归特征消除 基本选择方法+相关性+树的重要性-管道 混合特征选择方法 功能改组 递归特征消除 递归特征添加
【文件预览】:
feature-selection-for-machine-learning-master
----11-Hybrid-methods()
--------11.6-Recursive-feature-addition-with-Feature-engine.ipynb(155KB)
--------11.4-Feature-shuffling-with-Feature-engine.ipynb(94KB)
--------11.1-Feature-shuffling.ipynb(33KB)
--------11.3-Recursive-feature-addition.ipynb(141KB)
--------11.5-Recursive-feature-elimination-with-Feature-engine.ipynb(155KB)
--------11.2-Recursive-feature-elimination.ipynb(141KB)
----.gitignore(97B)
----feature_selection.png(366KB)
----requirements.txt(205B)
----09-Embedded-Lasso()
--------09.2-Basic-methods-plus-Lasso-pipeline.ipynb(14KB)
--------09.1-Lasso.ipynb(23KB)
----04-Correlation()
--------04.1-Correlation-Pearson.ipynb(251KB)
--------04.3-Correlation-with-Feature-engine.ipynb(25KB)
--------04.4-Pipeline-with-Feature-engine.ipynb(8KB)
--------04.2-Basic-methods-plus-correlation-pipeline.ipynb(15KB)
----SAVE_DATASETS_HERE.txt(0B)
----07-Wrapper()
--------07.3-Exhaustive-feature-selection.ipynb(27KB)
--------07.1-Step-forward-feature-selection.ipynb(37KB)
--------07.2-Step-backward-feature-selection.ipynb(34KB)
----LICENSE(2KB)
----Prepare-Titanic-dataset.ipynb(9KB)
----trainindata.png(63KB)
----10-Embedded-tree-importance()
--------09.2-Random-Forest-with-recursive-feature-selection.ipynb(17KB)
--------09.3_Basic-methods-plus-Random-Forests-pipeline.ipynb(14KB)
--------09.1-Random-forest-importance.ipynb(32KB)
----README.md(2KB)
----08-Embedded-linear-coefficients()
--------08.3-Regression-coefficients-and-regularisation.ipynb(283KB)
--------08.2-Linear-Regression-coefficients.ipynb(16KB)
--------08.1-Logistic-regression-coefficients.ipynb(40KB)
--------08.4-Basic-methods-plus-coefficients-pipeline.ipynb(14KB)
----06-Filter-other-metrics()
--------06.3-Univariate-Performance-with-Feature-engine.ipynb(140KB)
--------06.1-Univariate-roc-auc.ipynb(77KB)
--------06.2-Method-used-in-a-KDD-competition.ipynb(32KB)
--------06.4-KDD-method-with-Feature-engine.ipynb(14KB)
----05-Filter-Statistical-Tests()
--------05.4-Basic-methods-plus-statistical-pipeline.ipynb(17KB)
--------05.2-Fisher-score.ipynb(10KB)
--------05.1-Mutual-information.ipynb(82KB)
--------05.3-Univariate-selection.ipynb(79KB)
----03-Constant-Quasi-Constant-Duplicates()
--------03.4-Constant-features-with-Feature-engine.ipynb(14KB)
--------03.2-Quasi-constant-features.ipynb(26KB)
--------03.3-Duplicated-features.ipynb(24KB)
--------03.5-Duplicated-features-with-Feature-engine.ipynb(12KB)
--------03.1-Constant-features.ipynb(23KB)