Styles & Signs Final Challenge

This is my Styles and Signs Final Challenge Project for the GWC Pathways Program!

Project Requirements

My project:

• Uses the correct (SignMNIST) training and test datasets.
• Changes the number and/or the split of the data between validation and training datasets.
• Changes the design of the neural net by changing the size of the hidden layer.

Attributions

*If you used any code, stories, or poems from another person or group of people, tell us about it here. Make sure it is in the public domain, has a license that allows you to use it, or is one of your own.

• Sign MNIST dataset source: https://www.openml.org/search?type=data&status=active&id=45082

---

File Overview

← README.md

README.md file give you more documentation and information about a program. They are super helpful for describing what a program should do, any issues you've encountered, changes you want to make, and more.

← main.py

This is where you will write your main program.

← ClassViewer.py

This utility program displays the first 10 images of the class class_number. You can run the program by typing python ClassViewer.py in the Shell tab.

← DataDumper.py

This utility program displays the first 5 and last 5 rows and the first two and last two columns of the SignMNIST data set. You can run the program by typing python ClassViewer.py in the Shell tab.

← DisplayClassHist.py

This utility program displays a histogram of the SignMNIST data set classes. You can run the program by typing python DisplayClassHist.py in the Shell tab.

← ShowPixelGrid.py

This utility program displays a one image from the SignMNIST data set classes with the numerical pixel values overlayed on each pixel. You can select which row of the dataset is displayed by changing the row number in the iloc location: row = df.iloc[7] You can run the program by typing python DisplayClassHist.py in the Shell tab.

← sign_mnist_13bal_test.csv

This is a comma separated (CSV) formatted file containing a test portion from a balanced sample of 13 images from each class in the SignMNIST dataset.

← sign_mnist_13bal_train.csv

This is a comma separated (CSV) formatted file containing a training portion a balanced sample of 13 images from each class in the SignMNIST dataset.