I would like to ask for your opinions regarding my machine learning model.

I have a 4000 dataset with 4 classes and 71 features. 3200 are for the training set and 800 is for the testing. I trained my model and achieved an accuracy of 60% on both training set and testing test. I also run the validation set producing a similar result. I used scikit learn's MLPClassifier to produce my model. I tweak all the parameters however, when I look at the training curve, it remains constant at 60% for both training and validation set when it hit the 1500th example. I cannot add more training data since I am limited with my options to add.