Boosting Machine Learning Algorithms: An Overview - KDnuggets

Combing various machine learning algorithms while solving a problem usually results in better results. The individual algorithms are referred to as weak learners. A weak learner is a model that gives better results than a random prediction in a classification problem or the mean in a regression problem. The final result from these algorithms is obtained by fitting them on the training data and combining their predictions. In classification, the combination is done by voting, while in regression, it's done via averaging.