In this repo I am looking at learning methods that will make my models better.
Today 5th Sept 2019 I have landed on something that was released this week called Neural Structure Learning.
Still trying to make sense of it all, but what I know is that they are incoporating graph theory, which can be derived from data instead of just features.
Leveraging structure signals helps make training better because we get higher accuracy.
I will improve what i know as time goes on.