Lecture 2: Regression and Regularization
Please Log In for full access to the web site.
Note that this link will take you to an external site (https://shimmer-mit-edu.ezproxy.canberra.edu.au) to authenticate, and then you will be redirected back to this page.
Lecture Feedback Form | | Slides in PDF
Interactive Slides:
Two common questions about the recording:
1. At about 3:40 the robot video through HDMI drowned the mic sound out. What was said in class? (click to see transcription.)
What I said:Now, I just showed some "before" and "after" demos. I don't want the take-away message to be that 'oh, Machine learning is taking over the world and all old methods are dead.' No, not really.
Now, here, also about 10 years later, our lab-mates continued working on the old (optimization + control) methods on Atlas 2.0 here. See? Atlas is still doing well!! So, it's quite exciting that now everybody is improving, everybody is doing a better job. In fact, a bit of what goes on in controlling Atlas here borrowed some ideas from ML; and vise versa, the manipulation demo work I just showed also has ideas built on classical methods. Either way, hopefully for those of you who might continue on a research journey, you also get to enjoy this "inspired by each other" life style.
I then skipped the video to show Spot (the yellow dog bot) and Handle (the big ostrich-like bot). I commented a bit on how the dance scene was "quite a party" and and segued to the lecture party.
2. How do I switch between video streams or camera angles? (click to see how-to.)
Recording