You will find the solution of the Poker Hand Dataset here. This dataset is having a stragihtforward way of solving the problem. You don't need to impute data, feature scaling or any other kind of data processing apart from Encoding data so that it is meaningful to model and you can have high accuracy.
We have acheived 91% accuracy rate using randomForestTree classifier.