Why not? If I understood correctly, you basically have 840 1D data points that come from two different distributions. You can fit a GMM on it, no issues.
I will take a look to see if I still have the code. Anyway, it may require a little bit of refactoring since it was quite messy. I will keep you tuned! :)
Thanks for the feedback! Maybe the video is a little bit confusing since I used a 1D equation while presenting a 2D visualization. Indeed, if you go beyond 1D, the variance becomes a matrix and you need to compute all the elements in it.
Thanks! Yes, you can take a look at the following explanation of multivariate normal distribution to get an idea of how the GMM formulas could extrapolate to multiple dimensions: ruclips.net/video/UVvuwv-ne1I/видео.html
Thanks for the suggestion! I added some subtitles, it may take a while for RUclips to process them and add the timings though. They may come in-handy for other viewers as well! :)
Video to the Multivariante Gaussian Distribution: ruclips.net/video/UVvuwv-ne1I/видео.html
I wrote my thesis on the application of GMM and I can confirm that this video is high quality.
I wrote my thesis on application of gaussian mixture models, and I still have no clue what they are lol
@@Cam-lh7nr 😆
Underrated video. You explained it so well and intuitively, thank you!
Many thanks! I am happy you found the explanation helpful! :)
I like the definition on the formula at 3:45. Everything is so clear this way. Thanks!!!!!!
Happy to hear that! :)
Thank you, I had a rough idea of the concept and this gives me a better foundation to understand topics based on it.
You're welcome! Happy to hear that this video helped you! :)
Perfect refresher! Very concise and understandable, thank you
Thanks! Glad it was helpful! :)
Very good and concise explanation, thank you!
Thanks! Glad it was helpful! :)
This was so clearly explained! 🌟
Thanks! Glad it was helpful! :)
Understandable video. Thanks for your sharing!
Thanks! Glad you liked it!
Short and precise. Thank you
Glad it was helpful! :)
Amazing explanation
Thanks for the feedback! I am happy you found the explanation helpful! :)
Superb amazing
You sort me out!
You're welcome! :)
can we still do this process if the first distribution is having 63 x 1 element and other distribition is 777x1 elements
Why not? If I understood correctly, you basically have 840 1D data points that come from two different distributions. You can fit a GMM on it, no issues.
how to use the GMM in the kalman filter?
Unfortunately, I'm not so familiar with the kalman filter to get into such details. :(
Can you release the code for making GMM demo animations?
I will take a look to see if I still have the code. Anyway, it may require a little bit of refactoring since it was quite messy. I will keep you tuned! :)
you need to calculte every element of the matrix. Why do you only mention the diagonal terms?
Thanks for the feedback! Maybe the video is a little bit confusing since I used a 1D equation while presenting a 2D visualization. Indeed, if you go beyond 1D, the variance becomes a matrix and you need to compute all the elements in it.
@@datamlistic how though? Do you have any source I can use? Nice video btw
Thanks! Yes, you can take a look at the following explanation of multivariate normal distribution to get an idea of how the GMM formulas could extrapolate to multiple dimensions: ruclips.net/video/UVvuwv-ne1I/видео.html
Please add the subtitles :(
Just look for another tutorial if you don’t understand, he’s pretty clear
Thanks for the suggestion! I added some subtitles, it may take a while for RUclips to process them and add the timings though. They may come in-handy for other viewers as well! :)