Homework 5 is released in three parts. To complete it, you must submit all parts. Here are the handouts:
- Deep Learning (80 points)
- MapReduce (90 points)
- Unsupervised Learning (100 points) - Only for students in the 15-688 version of the course.
Each part is a separate code lab in Diderot. You can (and should) submit each part as you complete it.
For Deep Learning, the usual 1 minute time limit on code execution has been increased to 15 minutes. For Unsupervised Learning, the usual 1 minute time limit on code execution has been increased to 6 minutes. Accordingly, we have imposed a limit on the number of submissions.
Experimentally, we have included Python type notations in MapReduce. These are optional to use and understand.
We will hold a recitation (see schedule) for this homework in which we will review some basic concepts for this homework. Attempt all parts of the homework before attending the recitation to make the most of it.
The recitation notes are available here.