Summary
In this chapter, we understood the different types of prediction problems and how various algorithms aim to solve them. Then, we understood how the different ML algorithms are categorized into supervised, unsupervised, semi-supervised, and reinforcement based on their method of learning from data. Once we had an understanding of the overall problem domain of ML, we understood that H2O AutoML trains only supervised learning ML algorithms and can solve prediction problems in this domain specifically.
Then, we understood which algorithms H2O AutoML trains starting with GLM. To understand GLM, we understood what linear regression is and how it works and what assumptions about the normal distribution of data it has to make to be effective. With these basics in mind, we understood how GLM is generalized to be effective, even if these assumptions of linear regression are met, which is a common case in real life.
Then, we learned about DRF. To understand DRF, we understood what...
