In the top 1% of drivendata.org competition for predicting heart diseases out of 4000+ data scientists and data science teams. Achieved 0.3103 log loss, using deep neural network with sequential modeling.