An early approach on map difficulty estimation using Deep Neural Networks.
This is one of the first few “research” efforts I’ve done. It was one of the key motivators for me to try out a more research oriented role.
In hindsight, the report is rather hand-wavy, which as expected, won’t fly in
the real world. However, the key thing was that I had fun trying to do research!
Please consider the results below with doubts, it’s not a robust approach, it’s
just a quick college try on research 