Time delay embedding for Koopman

[@ogunstega7348]

Wow.. Very Interesting Lecture, I learned and I also smiled when learning..

[@qilinwang5889]

Thank you for this video, it is so helpful!

[@hamidrezakarbasian5786]

Hi, it is really fruitful. But I found that it is difficult to track each concept, while I dont know which video is related to which one. It would be great if you categorize the course using play-list. Thanks for lectures!

[@zoloojagaa9198]

faculty.washington.edu/kutz/am563/page1/page15/am563.html I think you can see the playlist here

[@stanleybak1363]

The good time delay was 9 seconds, which is more than the time it takes to complete a single cycle in vanderpol. In this case, wouldn't the future state prediction essentially end up just looking at what the system was doing at the same point in the previous cycle, since that's a state variable that's available? Also: is there a trick to start the linearized simulation? With extended DMD you could multiply the initial state to get the values of the observations, whereas now it looks like you need to know the historic state to construct the initial vector.

[@adamgerlach6379]

Dr. Kutz, in the video you mentioned 'in the last video/lecture' a couple of times. What video are you referring to? May I suggest putting playlists together to help organize the videos? Thanks for the great lectures.

[@zoloojagaa9198]

faculty.washington.edu/kutz/am563/page1/page15/am563.html I think you can find the playlist here

[@RonDesGroseilliersJr]

Can you apply this to the financial markets?

[@AliRashidi97]

Wow Just wow

[@gegebao5363]

omg