Fashion MNIST Neural Net. Classifier

o o In this report and research project, Guiragos et al. document the culmination of their artificial neural-network architecture exploration. The primary focus was the implementation of an efficient and accurate classifier of the fashion-MNIST database through an expanded multilayer perceptron neural network with backpropagation. To rephrase, Guiragos et al. use the MATLAB coding language to create a machine-learning algorithm which can accurately classify images of clothes in the fashion-MNIST database. Overall, wish to demonstrate our thorough understanding of our learning algorithm as it is expanded by a variety of optimization techniques.

o The project serves as a synthesis of ANN course knowledge and current academic literature, showing ability to understand and apply improvements to our work which come from outside of the course content. Further, it is the culmination of my data science work in IAS, my mathematics work in gen-ed, and my programming ability developed in CSS.