Goal of the project is to use Gradient Descent for logistic regression from scratch to train a model using a group of hyper-parameters to classify whether a patient has diabetes(class 1) or not (class 0). Based on Pima Indians diabetes dataset.
Then Neural networks model is trained on the same dataset to compare the output of the previous model.
- Achieved an accuracy of 77.124% implementing Stochastic Gradient Descent from scratch.
- Achived an accuracy of 78.512% training a Neural Network
