In my work and following the lessons of the course, I sometimes want to try applying different methods to solving machine learning problems on common datasets. e.g, correctly categorizing the well known Reuters article data set which is also available in keras.
however, It’s kind of hard to find out what are the state of the art results for this (and other) problems, so I don’t really know how good are my results, or perhaps the problem was already solved to my satisfaction, and I should move on to a different problem.
I find that going through random articles might by frustrating and not fruitful. what is the correct way to address this in your opinion?