a Japanese introduction of an R package {featuretweakR }
available from: https://github.com/katokohaku/featureTweakR
reference: "Interpretable Predictions of Tree-based Ensembles via Actionable Feature Tweaking" (https://arxiv.org/abs/1706.06691). Codes are at my Github (https://github.com/katokohaku/feature_tweaking)
11. 欠損値の補完
missForestによる欠損値補完 in “Imputation of Missing Values using Random Forest”
https://www.slideshare.net/kato_kohaku/imputation-of-missing-values-using-random-forest
@TokyoR#53
ちょっと変わった使い方...
12. ルール抽出・要約
ランダムフォレストにバスケット分析 in “Interpreting Tree Ensembles with inTrees”
https://www.slideshare.net/kato_kohaku/interpreting-tree-ensembles-with-intrees
@TokyoR#51
defragTreesも良い
...が、R実装がない
ちょっと変わった使い方...
14. 利用者に納得感を与える変数選択法
LASSOの別解を与える in “Introduction of "the alternate features search" using R”
https://www.slideshare.net/kato_kohaku/introduction-alternate-featuresinlassor-71186764
@TokyoR#58
例えば...
36. step-by-step procedure
1. Installation
2. data preparation
3. exploring randomForest
1. build randomForest
2. model shrinkage (feature selection) based on importance
3. scaling feature-selected data
4. performance comparison forest with all-feature v.s. selected-
features
4. Step-by-step procedure
1. extract rules
2. set modified rules (e-satisfactory instances)
3. predict individual suggestion for each instance
5. restore suggestion from scaled feature to original scale.
6. Visualize suggestion
FEATURE TWEAKING
To follow step-by-step procedure, please see:
• https://github.com/katokohaku/featureTweakR/blob/master/README.Rmd
37. build randomForest
Step-by-step
• To view variable importance and number of trees required.
To follow step-by-step procedure, please see:
• https://github.com/katokohaku/featureTweakR/blob/master/README.Rmd
38. build randomForest
Step-by-step
• To view variable importances and number of trees required.
To follow step-by-step procedure, please see:
• https://github.com/katokohaku/featureTweakR/blob/master/README.Rmd