Glossary Terms

Backpropagation

Backpropagation – A training process where errors are propagated backward to update the model’s parameters..

Bias

Bias – Systematic errors in model predictions due to data imbalance or other factors; also refers to adjustable parameters.

Bigram

Bigram – A sequence of two words or characters, used in text processing.

Binary Classification

Binary Classification – Classification tasks with two possible categories or outcomes.