DPR-MNIST challenge
This challenge was part of the course in 'Detection and Pattern Recognition' at the University of Stuttgart. The task was to develop a classifier in Matlab to reach maximum accuracy on the MNIST dataset.
The convolutional neural network implemented reached 99.56% accuracy and used data augmentation.
File containing the main script: https://github.com/MalteEbner/DPR_MNIST/blob/master/NN/deepNN.m