Welcome to the Intro to Machine Learning Workshop for TrojanHacks Spring 2018

Please read the follwing instructions carefully before the workshop

Overview

This workshop will focus on deep learning. We will go over some basic concepts about deep learning and then build a classifier on the hand-written digit dataset (MNIST) with tensorflow.

Environment Setup

An instruction on how to setting up the environment could be fond here here . Please note that you only have to install tensorflow, numpy and matplotlib for this workshop. Therefore, after completing step 7 in the instruction, simply skip step 8. In step 10, you do NOT need to run these two commands: pip install pandas and pip install keras. Pandas and Keras are also two very popular packages in machine learning. However, we will not cover these two libraries in this workshop.

After completing the above steps, you should be able to run the following lines of codes

import matplotlib.pyplot as plt
import numpy as np

import tensorflow as tf
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)
plt.imshow(mnist.train.images.reshape(mnist.train.labels.shape[0],28,28)[0].reshape(28,28), cmap='gray')
plt.show()

This block of codes will load the mnist data set and will output a black and white image of the digit 7, which is the first image in the training set.

Clone this repo into your own machine

You should then clone this repo into your own machine. In this repo you can find two additional files demo_incomplete.py and demo_incomplete.ipynb. These are the skeleton code for the workshop. And we will be building on this to complete our classifier.