data_analysis

For Data Mining

0. EDA

Visualize

Feature Importance(gain, split)/Feature_Permutation/Partial Dependence Plot

correlation

Adversarial Distribusion

1. Preprocessing

Outlier

-Construciton

Impute

-Impute by BaseAggregation

-Impute by Regression(LGBM)

Replace

2. Feature Engineering

Categorical Encoding

-Likelihood Encoding(Target Encoding)

-Count Encoding

-One Hot Encoding

-Label Encoding

Base Aggregation

Dimentional Reduction(Embedding)

-PCA

-LDA

-tSNE

-UMAP

Clustering

-Kmeans

-EM Algorithm

3. Feature Selection

Feature Importance (LGBM)

Feature Permutation

4. Feature Management

5. Various Machine Learning Algorithm

6. Explainer

Shap

LIME(Unexecuted)

99. Important