AI Dictionary › AI Fundamentals

Random forest

Random forest is a machine learning model that combines the predictions of many different decision trees to achieve a more accurate and stable result than a single tree would produce. The underlying idea is that a large group of imperfect but diverse models, voting together, makes fewer mistakes than any single model, however well built.

Definition

How it works

Each tree in the forest is trained on a random sample of the data, drawn with replacement from the original dataset, and at each of its nodes it can consider only a random subset of the available features instead of all of them. This double injection of randomness makes the trees different from one another, preventing them from all making the same mistakes. The final prediction comes from having the trees vote in classification, or averaging their outputs in regression.

Applications

It is one of the most widely used models in production for tabular problems, data organized in rows and columns like a spreadsheet or database: fraud detection, customer churn prediction, insurance risk assessment, medical diagnosis based on lab tests. It delivers good performance without much tuning effort, holds up well against noisy data, and also provides an estimate of which features matter most in the decision.

History & etymology

The term random forest, literally a random forest, plays on the fact that many decision trees together form exactly that, a forest, and that the randomness in choosing data and features is what makes them collectively stronger. The algorithm took its definitive shape with the work of Leo Breiman, a statistician at the University of California, Berkeley, who published its finished formulation in a 2001 paper, building on earlier ideas of bagging and random feature selection developed in the second half of the 1990s.

Related terms

• Decision tree
• Gradient boosting
• Ensemble learning
• Feature importance
• Overfitting

More in AI Fundamentals

• LLM (Large Language Model)
• Token
• Context Window
• Temperature
• AI Hallucination
• Fine-tuning
• RAG (Retrieval Augmented Generation)
• Embedding

Put it into practice

• Real scenario: apply this concept with an instant AI verdict.
• Daily gym: short drills to train dictionary concepts.
• Playoff game: challenge other players, elimination-style.

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.