Dog Breed Classification Using Convolutional Neural Network

Experimented and built a convolutional neural network using Keras to identify a particular dog breed from amongst 133 categories with a data set of 8351 pictures of dog species

The achieved accuracy of the neural network is 3% with 10 epochs

Analyzed different CNN architectures such as VGG16, VGG19 and Resnet to improve the current neural network performance

Part of the Udacity Machine Learning Engineer Nanodegree

To see the project open dog_app.ipynb file