Random forests
Era 3 · From the expert boom to the statistical turn (1980–2011) › Statistical learning and probabilistic models
- Lineage
- Data and learning
- Kind
- Publication
Who
AT&T Bell Labs, University of California, Berkeley
What happened
In 1995 Tin Kam Ho of Bell Labs presents random decision forests at a document analysis conference. In 2001 the statistician Leo Breiman of Berkeley gives the method its definitive form in "Random Forests."
Why it mattered
Many decision trees, each trained on a random part of the data or the variables, vote together. The ensemble corrects each tree's overfitting, and it is one of the most widely used methods for tabular data.
The full story
A decision tree classifies by asking successive questions about the variables: whether age is above some value, whether income is below another. It is easy to understand, but a deep tree fits its training data too closely and generalizes poorly.
Ho proposed building many trees, each restricted to a randomly chosen subset of the variables, and combining their answers by vote. Since each tree sees the problem from a different angle, their errors tend to cancel out, and the forest can grow without overfitting. In 1997 Yali Amit and Donald Geman explored a similar idea, choosing at random among possible splits when building each node.
Breiman, who had already invented bagging, training each model on a random sample of the data, combined it with random selection of variables in the 2001 paper. That makes the trees less correlated with one another, and the ensemble more accurate. Adele Cutler worked with Breiman on developing the method.
Random forests became one of the most widely used methods in machine learning, especially with data organized in tables.
Builds on
Learned decision trees (1984)
Documented direct influence
Breiman's random forests combine many decision trees built as in CART
Sources
Original source
Further reading
How to cite
“Random forests.” The History of AI, revised November 13, 2026. https://thehistoryofai.com/en/milestone/random-forests
This entry was last reviewed on November 13, 2026.