1. Idea
- Boosting is an ensemble learning technique that can generate a strong learner from a collection of weak learners.
- Two popular boosting algorithms are:
- A practical limitation of boosting is that it cannot be parallelized easily, because each model is dependent on the previous model.