When should we use a random forest?

The following are some examples of when to use random forests:

• When model interpretation is not the most important criterion. Interpretation will not be as easy as a single tree.
• When model accuracy is most important.
• When you want robust classification, regression, and feature selection analysis.
• To prevent overfitting.
• Image classification.
• Recommendation engines.