layout: default title: 18.065 parent: Spring 2023 nav_order: 1 —
‘18.065 Matrix Methods in Data Analysis, Signal Processing, and Machine Learning’
Table of contents
- Course Info
- Realistic Prerequisites
- Subject Matter
- Course Staff
- Lectures
- Problem Sets
- Exams
- Resources
- Grading
- Advice to Future Students
Course Info
Class Size | 70 |
Hours/Week | 11.2 (28 responses) |
Instructors | Steven Johnson |
Overall Rating | 6.2/7.0 |
Realistic Prerequisites
- People recommend having taken 18.06 as a hard prerequisite.
- The class does review linear algebra in the beginning but at a faster pace.
Subject Matter
- Students described the content as useful, foundational, broad, practical, with a mix of theoretical and applied content.
- The ML part was covered in depth in certain areas.
Course Staff
- There were no TAs/LAs, but the professor was helpful, approachable, and enthusiastic in office hours, and gave more detailed answers in office hours compared to Piazza.
Lectures
- The professor mostly wrote notes as he taught, sometimes having demo code and slides.
- The students thought the lectures were the most helpful and the most referenced resource.
- Some students also found the textbook helpful.
Problem Sets
- There were biweekly problem sets which students spent 4-10 hours on.
- Students thought the problems were relatively straightforward and canonical, though some students thought they were also interesting and fun.
Exams
- There were no exams, but a final project.
- One student thought that the final project was the best part of the class.
Resources
- The professor posted handwritten notes to canvas, and people found the textbook very helpful.
- There was an active Piazza.
- The textbook was less relevant later in the semester.
Grading
- Students felt that the grading was lenient but slow.
Advice to Future Students
- ”Brush up on linear algebra skills if it has been a while since taking a course.”
- “Find a good group for the psets otherwise you will give up.”