I've used various machine learning algorithms to create more informative projects about diabetes.These are Logistics regression, Decision tree, Random forest, XGB Classifier, GBM Classifier, and the EXT Classifier. These give us great accuracy from 90% to 93%.
XG Boost performed the best with accuracy of 93%
Google Colab: https://colab.research.google.com/drive/1YSQgVXrRh-x7xVBk66LQWxhmdY1PgJat?authuser=1