The purpose of this project was to gain experience with Tensorflow and Keras by working with a diabetes dataset found in the diabetes.csv file. The data was already tidy so most preprocessing steps were bypassed. Multiple neural networks were created to determine how well they predicted whether a person has diabetes or not. They were evaluated through multiple metrics including accuracy, precision, recall, and F1 score.
Depending on how large or small the neural network was, it performed differently in all cases. Some structures performed better than other types of structures. This could be seen through the plotted graphs.
This was my first project involving Tensorflow and Keras, and I think it turned out well. Each neural network made predictions with reasonable accuracies. I enjoyed working with this project, and I am sure I will continue to work with neural networks in future projects.