Random Forest

A random forest is a machine learning method that builds many decision trees, each trained on a random sample of the rows (bootstrap sampling) and considering only a random subset of features at each split. For classification the trees vote; for regression their predictions are averaged.

Why it matters

A single tree overfits easily, but averaging many diverse trees cancels much of their individual noise. Random forests often deliver strong accuracy on tabular data with little tuning, handle missing and mixed features reasonably well and give a rough measure of feature importance. They are less interpretable than one tree, larger in memory and slower to predict than simple models.

Example

Train 200 trees on a churn dataset.
New customer -> 140 trees say "will churn", 60 say "stay"
Prediction: churn (70% of votes)

The vote share is a rough confidence signal, although it is not a calibrated probability by default.

Common mistakes

• Treating feature importance as proof of cause or trusting it when features are strongly correlated.
• Adding trees endlessly. Returns diminish after a point.
• Leaking future information into features, which inflates validation scores.

Related

• Decision tree
• Overfitting
• Classification
• Machine learning

यह पेज सामान्य जानकारी है, जिसे हमारी संपादकीय टीम ने लिखा है और दस्तावेज़ीकरण बदलने पर अपडेट किया जाता है। सर्वर और प्रदाता अलग-अलग होते हैं, इसलिए बदलाव पहले सुरक्षित तरीके से जाँचें। देखें: संपादकीय नीति और अस्वीकरण.