feat: preprocessing transformation priorities

[jitingxu1]

Building upon the deliverables outlined in issue #19, the objective is to enhance the coverage of ibisml machine learning preprocessing transformations, prioritizing key areas for improvement.

Please share your favorite ML transformation for your daily ML tasks and provide additional context as to why you find it particularly useful.

Assumption

• Raw feature creation is done using ibis
• tabular data

Priority definition:

• P0: Essential tasks vital for the model development, Essential before our initial release.

• P1: Desirable tasks that can enhance the model, based on feedback and further optimization.

• P2: Additional tasks aimed at improving the model, based on feedback and further optimization.

Priorities

Preprocessing Module	Ibis-ml Step	sklearn	Priority	Status	Note	Model Needed
Encoding	CatgoricalEncode	OrdinalEncoder	P0	Done
Encoding	CountEncode		P1	Done
Feature Engineering	CreatePolynomialFeatures	PolynomialFeatures	P0	Done
Non-linear Transformation	Math Transformation (log, sqrt,)		P1	Done	ibis
Standardization and Scaling	ScaleStandard	StandardScaler	P0	Done		KNN, MLPBased, SVM
Encoding	TargetEncode	TargetEncoder	P0	Done
Feature Reduction	DropZeroVariance	VarianceThreshold	P0	Done
Imputing	HandleUnivariateOutliers	SimpleImputer	P0	Done
Feature Engineering	ratio variable creation		P0	Done	ibis
Discretition	DiscretizeKBins	KBinsDiscretizer	P0	Done
Discretition	Feature binarization	Binarizer	P1	Done
Standardization and Scaling	ScaleMinMax	MinMaxScaler	P0	Done		KNN, MLPBased, SVM
Custom Transformer	Custom transform	FunctionTransformer	P0	Done
Encoding	OneHotEncode	OneHotEncoder	P0	Done
Imputing	Outlier - Impute and capping		P0	Done		Log/Linear Reg
Feature Reduction	Continuous Target Mutual Info		P1	Not started
Feature Reduction	Discrete Target Mutual information		P1	Not started
Feature Engineering - Text	Count Transfomer	CountVectorizer	P2	Not started
Feature Engineering - Text	TFIDF Transformer	TfidfTransformer	P2	Not started
Encoding	label binarizer	LabelBinarizer	P2	Not started
Encoding	label encode	LabelEncoder	P2	Not started
Standardization and Scaling	MaxAbsScaler	MaxAbsScaler	P2	Not started
Standardization and Scaling	RobustScaler	RobustScaler	P1	Not started		KNN, MLPBased, SVM
Imputing	Missing value - Nearest Neighbor	KNNImputer	P1	Not started	Doable
Non-linear Transformation	QuantileTransformer	QuantileTransformer	P1	Not started
Non-linear Transformation	Inverse and Logit transformation		P2	Not started
Imputing	Missing value - Linear reg		P1	Not started	Not Support
Imputing	Missing value - bagged trees		P1	Not started	Not Support
Feature Reduction	Filter col with missing rate threshold		P1	Not started
Feature Reduction	Filter Feature by high correlation		P2	Not started	Doable
Non-linear Transformation	PowerTransformer	PowerTransformer	P1	Not started		MLPBased, SVM
Feature Reduction	PCA		P1	Not started	Not Support
Imputing	Missing Value - rolling window Imputing		P2	Not started
Feature Engineering	Spline transformer	SplineTransformer	P1	Not started

Reference:

• Preprocessing methods for different models
• Tidymodels: Recipe steps
• Sklearn preprocess
• Tidymodels: Ordering of preprocessing