DATA 442: Assignments

Please note the below may change throughout the course. Any changes will be announced on Piazza. All assignments and assignment materials can be found below. You must upload your modified submission.py to Gradescope (www.gradescope.com) before 11:59PM on the date an assignment is due. No late submissions are accepted.

The first time you sign up for Gradescope, you will need an entry code to access this course. For Spring 2021, the code is MZ2R7Y. Please make sure to sign up for Gradescope using your William & Mary credentials; otherwise, I may be unable to link your name back to W&M systems.

Most lab files are provided in .tar.gz format. If you are unfamiliar with this format, 7zip will be useful! Once you open each lab, you will find numbered files. Start with the file 1-README.md and work your way through. Once you've finished all of the numbered excersizes, you should be ready to complete submission.py.

DATA 442 Bug Bounty Program: If you find an error in documentation, code, or this website, you can earn extra credit on your final grade. To be eligible for points, you must provide the following information:

Bug bounties should be emailed directly to the instructor (danr@wm.edu). Bounties will be awarded on a first-come-first-serve basis; two individuals cannot get credit for the same bug. If your solution is accepted, you will be awarded bonus points on a sliding scale from 0 to 10 points, depending on the severity of the issue identified and the completeness of your solution. One misspelled word will not get you a point, but a bunch reported at once might :).
Title Description Due Date Materials
Lab 0: Making Sure You Understand How You Will Be Graded In this example assignment, you must code a calculator with functions which multiply and add. For a python refresher, this is a great resource. February 8th
5 bonus points 		Lab 0 Download
Lab 1: Baseline Computer Vision This assignment covers data manipulation, test/train splits with image data, KNN and Linear classification of linear data, and optimization. A backup URL for the CIFAR data can be found here. March 5th
20% of Grade
Max Score: 110
(100 + 10 Bonus) 		Lab 1 Download
Midterm Once started, you will have 45 minutes to complete the midterm (you cannot pause the timer). The midterm has 7 questions (most of which have sub-questions),and is taken on gradescope. Points will not be given for technical issues. It is recommended you take the midterm at a location (i.e., the library) with multiple computers you can access. It is up to you to ensure you have a ’plan B’ in case of technical issues with your computer. The midterm must be taken independently (i.e., no groups), but you can use any notes or other resources you would like. The midterm covers materials up to and including this week - make sure you watch lecture 7 before starting! You must complete the midterm by:
March 12th, 11:59PM
20% of Grade
Max Score: 100
 Gradescope
Lab 2: Neural Networks This lab teaches the fundamentals of neural networks, including back and forward propagation, loss function implementation, weights initialization, convolutional and affine layers, and more. A backup URL for the CIFAR data can be found here. April 9th
20% of Grade
Max Score: 100
 Lab 2 Download
Lab 3: Operational Networks This lab goes into more depth for networks, using common tools (i.e., Keras) to implement a range of network architectures. You will implement four "real" networks, building your own datasets and models to test against unseen data. May 7th
20% of Grade
Max Score: 100
 Lab 3 Download
Final Once started, you will have 45 minutes to complete the midterm (you cannot pause the timer). The final is the same format as the midterm, and is taken on gradescope. Points will not be given for technical issues. It is recommended you take the final at a location (i.e., the library) with multiple computers you can access. It is up to you to ensure you have a ’plan B’ in case of technical issues with your computer. The final must be taken independently (i.e., no groups), but you can use any notes or other resources you would like. The final covers material from the entire course, inclusive of the first half. You must complete the final by:
May 18th, 11:59PM
20% of Grade
Max Score: 100
 Gradescope