Summary
In this chapter, we discussed the third out of four groups of hyperparameter tuning methods, called the heuristic search group. We discussed what the heuristic search method is in general and several variants of heuristic search methods, including SA, the GA method, PSO, and PBT. We saw what makes each of the variants differ from each other, along with the pros and cons of each. At this point, you should be able to explain heuristic search in confidence when someone asks you. You should also be able to debug and set up the most suitable configuration of the chosen method that suits your specific problem definition.
In the next chapter, we will start discussing multi-fidelity optimization, the last group of hyperparameter tuning methods. The goal of the next chapter is similar to this one’s: to provide a better understanding of the methods that belong to the multi-fidelity optimization group so that you can explain those methods in confidence when someone asks you...