A small MLP programme from Kaggle:https://www.kaggle.com/competitions/playground-series-s4e10
Some fun experiments i did for MLP, without doing anything to filter the data, it already gives back 93 percent correct rate, better than XGboost, LightGBM, Adaboost.