Random Forests (RFs) can do classification or regression modeling.

Random Forests are one of the workhorse of modern machinelearning. Especially, they cannot over-fit to the training set, arefast to train, predict fast, parallelize well and give you a reasonablemodel even without optimizing the model's default hyper-parameters. Inother words, it is hard to shoot yourself in the foot while trainingor exploiting a Random Forests model. In comparison, with deep neuralnetworks it is very easy to shoot yourself in the foot.

Using out of bag (OOB) samples, you can even get an idea of a RFsperformance, without the need for a held out (test) data-set.

Their only drawback is that RFs, being an ensemble model, cannot predictvalues which are outside of the training set range of values (this isa serious limitation in case you are trying to optimize or minimizesomething in order to discover outliers, compared to your trainingset samples).

For the moment, this implementation only consider a sparse vectorof integers as features. i.e. categorical variables will need to beone-hot-encoded.For classification, the dependent variable must be an integer(encoding a class label).For regression, the dependent variable must be a float.

Bibliography

Breiman, Leo. (1996). Bagging Predictors. Machine learning, 24(2),123-140.

Breiman, Leo. (2001). Random Forests. Machine learning, 45(1), 5-32.

Geurts, P., Ernst, D., & Wehenkel, L. (2006). Extremely RandomizedTrees. Machine learning, 63(1), 3-42.