Textbook of AI

1943–, Computer scientist

Also known as: J. Ross Quinlan

John Ross Quinlan is an Australian computer scientist whose ID3 algorithm (Iterative Dichotomiser 3, 1986) introduced information gain as a splitting criterion for decision-tree induction. His 1993 successor C4.5 added handling of continuous attributes, missing values and pruning to combat overfitting; the book C4.5: Programs for Machine Learning came with full source code and made C4.5 the canonical decision-tree algorithm for a generation. C5.0 (1997 onwards) extended C4.5 with boosting and rule-extraction.

Quinlan's algorithms exemplify the pre-deep-learning machine-learning paradigm of interpretable, fast, low-tuning-overhead supervised learning. Decision trees remain a central building block of modern boosted-tree ensembles (XGBoost, LightGBM) which dominate tabular machine-learning competitions. Quinlan founded RuleQuest Research and remains active.