text Topics sentense1 [A,B,C,D] sentence2 [dhd,B,W,A,Q,G,WE,ffjeij,as] sentence3 [Z,AA,RF,Q,H,P,H,A,Y]
I have 4M rows in training set and actually, the sentence is an Article
Total Number of Topics we have approx 10k
Now for a given article, I have to predict topics. Number of topics can vary from 0 to 10k
I am new in ML, I am not able to think of how multilabel classification can handle 10K topics. please guide to solve this problem.
I am mainly interested in accuracy.