- Photo
- —
- Name
- —
- Authors
- Leo Breiman
- Year
- 2001
- Field
- Machine learning / Ensemble methods
- One-line gist
- Shows how averaging lots of noisy decision trees with randomness baked in gives you a shockingly strong predictor.
- Why I care
- Any time I need a baseline model that punches above its weight, this is still the first jersey off the bench.
- Best stat or figure
- The error vs. number-of-trees plots that flatten out like a good defensive rotation—diminishing returns, but steady.
- Difficulty
- Medium; readable for practitioners, the theory parts get spicy but skimmable.
- Where to read
- Machine Learning journal or the author’s reprints page online.
- My take
- It’s the paper behind half the Kaggle gold medals and more than a few production systems nobody brags about but everyone trusts.
TheLysts