I am currently looking research ideas focusing on where I could go as a master’s student. It seems as there is a lot of recent towards reinforcement learning in Deep Learning to Learn, and algorithms which explore this such as SNAIL, MAML, etc. I’m currently brainstorming ideas that I could potentially explore in this area as further research or study, but something practical enough to tackle in the course of a year. Some ideas I had were studying various algorithms which explore robots learning to learn an environment that is constantly changing in real time (this could cover domain randomization, simulation training, etc.). Or perhaps there are deep learning to learn techniques that I can explore through Content Based Image annotation and clustering algorithms such as affinity propagation. Or perhaps there might be something I am not exploring and should when it comes to deep learning, image learning, environment learning, etc.

More Christopher Klecker's questions See All
Similar questions and discussions