Introduction to Supervised Machine Learning

October 2018 – February 2019

I started with Kaggle to aquire datasets which I will use to create and test my neural network. My completed my first challenge called Digit Recognizer where I created a neural network that can identify handwritten digits. I used the Tensorflow library to create my neural network.

Next, I worked on classfing images. I wanted to specifically train my neural network on my dedicated Nvidia graphics card because it is much faster. In order to do that, I had to set up Nvidia CUDA, a parallel computing platform that allows developers to use their Nvidia graphics card for general purpose processing. I coded in Python using Juypter Notebook. I used the Keras Library to create a neural network that can classify images of dogs and cats. The dataset was found in the Dogs vs. Cats challenge which I completed successfully.

Dog Images Dataset Cat Images Dataset Training Neural Network on my GPU
Dog Images Dataset Photo Cat Images Dataset Photo Training Neural Network Photo