Introduction Lecture - Thurs, Aug 29
Head GSIs (Sam Weismann and Sarika Madhvapathy) can be reached at firstname.lastname@example.org
- Course website is available here.
- Use Piazza and Gradescope.
- Discussion at 1PM in Barrows 60.
- No prior experience in linear algebra.
- Will host a couple of round table events.
- Signups will be released in the coming few weeks.
- Due Friday at midnight
- Except HW0, due Wednesday, Sep 4 at 11:59 PM
- HW Party: Wed 2-4 PM and Thu 9-11 AM and 2-4 PM
- Mostly submitted using Gradescope
- Select pages
- Submit iPython sheets
This class is not curved. Competition is discouraged.
Key Ideas // Buzz Words
- Feedback, eigenvalues, stability
- Least-squares (ahem Data 8)
EE 16A focuses on the first 3 modules of the series:
- Module 1: Introduction to Systems
- Module 2: Introduction to Circuits and Design
- How do we use a model to solve a problem?
- Module 3: Introduction to Machine Learning
- How do we "learn" models from data?
EE 16B focuses on the next 3 modules of the series:
- Module 4: Advanced Circuit Design
- Module 5: Introduction to Robotics
- Module 6: Introduction to Unsupervised Learning
See photo library for chart of where to go after EECS 16AB.
Deisgn of Information Devices and Systems
- 16A focuses on how to use technology to sense the environment and process this information.
- 16B then adds to this with actuation -- what to do with this processed information.
- Best when hardware and software work together
- Understanding sensing and computing mechanisms is key to writing the best algorithms and code.
- Understanding the physical limitations of a situation is key to designing the best devices.
Module 1: Imaging
- Medical imagery of 1632 involved physical surgery to see what was happening inside.
- Modern-day imagery involves things like MRIs, CT scans, X-Rays, and Ultrasound.
- This (tomography) is the foundation of the first module.
How does tomography 'see' inside?
- Magnetic fields interact differently with different parts of your body.
- These different types of interactions can be used to figure out what's going on.
- A single measurement is not enough. Tomography will take many measurements (also called projections) from various angles.
- When combined, these will produce the image you're trying to uncover.
Fundamentally, this involves linear algebra.
Given an image with measurements:
- What do pixel values represent?
- density, population, etc.
- Can we solve for the values from projections?
There would usually be as many equations as there are unknowns, just as in past classes.