What is probabilistic classification and how is it different from hard classification?

Probabilistic classification outputs class probabilities, while hard classification outputs final labels only.

Probabilistic classifiers estimate P(Y|X), providing confidence scores for each class. Hard classification directly outputs the most likely class. Probabilistic outputs allow threshold tuning, ranking, and risk-aware decisions. Logistic regression and Naive Bayes are probabilistic, while decision trees can be used in both modes.