Expand description
Data encoding and categorical feature transformation utilities
This module provides comprehensive data encoding implementations including label encoding, one-hot encoding, ordinal encoding, binary encoding, hash encoding, frequency encoding, target encoding, feature hashing, categorical transformations, cardinality reduction, embedding-based encoding, statistical encoding, smoothing techniques, regularization methods, cross-validation encoding, time-aware encoding, and high-performance categorical feature processing pipelines. All algorithms have been refactored into focused modules for better maintainability and comply with SciRS2 Policy.
Structs§
- Binary
Encoder - Binary encoder for high-cardinality categorical features
- Binary
Encoder Config - Configuration for BinaryEncoder
- Binary
Encoder Fitted - Fitted state of BinaryEncoder
- Categorical
Embedding - Categorical embedding using neural network-style embeddings
- Categorical
Embedding Config - Configuration for CategoricalEmbedding
- Frequency
Encoder - Frequency encoder transforms categories to their occurrence frequencies
- Frequency
Encoder Config - Frequency encoder configuration
- Hash
Encoder - Hash encoder for categorical features using feature hashing
- Hash
Encoder Config - Configuration for HashEncoder
- Label
Encoder - Label encoder for transforming categorical labels to integers
- OneHot
Encoder - One-hot encoder for categorical features
- Ordinal
Encoder - Ordinal encoder for categorical features with inherent ordering
- Target
Encoder - Target encoder using target statistics for categorical encoding
Enums§
- Hash
Method - Hash function options
- Rare
Strategy - Strategy for handling rare categories
- Unknown
Strategy - Strategy for handling unknown categories