No handling Dummy encoding Target encoding Ordinal encoding Frequency encoding Presence flag Hashing Custom , drop missing , impute missing Presence flag Avg-std rescaling Min-max rescaling Binarization Custom Binning in {{proc.quantile_bin_nb_bins}} quantiles Cyclical datetime encoding Term hashing Term hashing + SVD TF/IDF vectorization Count vectorization Text embedding Custom Unfolding Custom Image embedding

Control value:

Preferred class:

Encoding map


	{{encodingValue\|number:3}}

Categorical feature coding

Encoding a categorical feature consists in replacing each value of the feature by a numerical value. There are different methods to compute this numerical value, including target encoding (impact, GLMM), ordinal encoding, and frequency encoding.

Learn more

Count vectorization

Used words: {{modelData.preprocessingReport.countvec[proc.name].used_words}}
Dropped words: {{modelData.preprocessingReport.countvec[proc.name].dropped_words}}

TF/IDF vectorization

Used words: {{modelData.preprocessingReport.tfidfvec[proc.name].used_words}}
Dropped words: {{modelData.preprocessingReport.tfidfvec[proc.name].dropped_words}}

Text embedding

Used model: {{sentenceEmbModel.friendlyName}}
Max tokens limit: {{sentenceEmbModel.tokensLimit}}

Image embedding

Used model: {{imageModel.friendlyName}}

Input features