[SOLVED] ECE472-Homework3

Description

Classify mnist digits with a (optionally convoultional) neural network. Get 005         at least 95.5% accuracy on the test test.

008          Problem Statement      Consider the mnist dataset consisting of 50,000 training

009

images, and 10,000 test images. Each instance is a 28 28 pixel handwritten digit

010                                                                                                                              ×

011          zero through nine. Train a (optionally convolutional) neural network for

012        classification using the training set that achieves at least 95.5% accuracy on the test

013

set. Do not explicitly tune hyperparameters based on the test set performance, use

014

015             a validation set taken from the training set as discussed in class. Use dropout and

016              an L² penalty for regularization. Note: if you write a sufficiently general program

017 the next assignment will be very easy.

018

⁰¹⁹           Do not use the built in mnist data class from tensorflow.

020

021

022

Extra challenge (optional)     In addition to the above, the student with the fewest

023

024            number of parameters for a network that gets at least 80% accuracy on the test set

⁰²⁵         will receive a prize. There will be an extra prize if any one can achieve 80% on the

026

test set with a single digit number of parameters. For this extra challenge you can

027

028          make your network have any crazy kind of topology you’d like, it just needs to be ⁰²⁹          optimized by a gradient based algorithm.