search

CodingTrain / Suggestion-Box

A repo to track ideas for topics
571 stars 86 forks source link

k-armed bandits #1391

Open kwichmann opened 4 years ago

kwichmann commented 4 years ago

Since reinforcement learning is a topic that is frequently suggested (and something I'm personally interested in), I think k-armed bandit would be a good introduction to the subject.

Exploration-exploitation tradeoff could be discussed. Greedy, epsilon-greedy, upper confidence bound etc. strategies could be implemented and compared.

shiffman commented 4 years ago

Thank you for this suggestion @kwichmann! I am really looking for good entry points so super appreciate this.

fahadmaqsood commented 4 years ago

@shiffman! you teach in a beginner-friendly way. I have learned a lot from your channel... Genetic Algorithm, Neural Networks, and Machine Learning... Your videos helped in getting started with all these. And it would be great if you teach Reinforcement Learning too.