That was great. I'm a network engineer and most of this is foreign to me. You made it simple to have a high-level understanding of different concepts. That's all I was looking for.
Great explanation of difference between vector and embedding, the meaning in ML/ML context and how they relate. Also, the journey from data to embedding is also really helpful. Thanks for sharing your thoughts!
Very helpful breakdown! I had heard the breakdowns used as "dimensions" but I appreciated you explaining how the "clusters" can catch vectors that fall within certain ranges or weights. Thank you!
I understand how this works for text, but I don’t get how it works for speech recognition. Are we comparing the similarity between pixels in the spectrogram?
Sorry for the late response, its this one for Chrome - chromewebstore.google.com/detail/super-simple-highlighter/hhlhjgianpocpoppaiihmlpgcoehlhio?pli=1
Points are representing data but the comprasion is made by vectors drawn between 2 data points. Comprassion and finding similarities are whole points of these databases.
That was great. I'm a network engineer and most of this is foreign to me. You made it simple to have a high-level understanding of different concepts. That's all I was looking for.
Great explanation of difference between vector and embedding, the meaning in ML/ML context and how they relate. Also, the journey from data to embedding is also really helpful. Thanks for sharing your thoughts!
Very helpful breakdown! I had heard the breakdowns used as "dimensions" but I appreciated you explaining how the "clusters" can catch vectors that fall within certain ranges or weights. Thank you!
This is extremely useful. Thank you so much Colin!
Very useful and concise. Helped me a lot!
I understand how this works for text, but I don’t get how it works for speech recognition. Are we comparing the similarity between pixels in the spectrogram?
what highlighter tool are you using?
Sorry for the late response, its this one for Chrome - chromewebstore.google.com/detail/super-simple-highlighter/hhlhjgianpocpoppaiihmlpgcoehlhio?pli=1
@@colintalkstech thx!
What software did you use to create the 3d colorful vector space? Is it MATLAB?
THe plots are from articles they are summarizing
Very interesting - thanks
Let's always do alot of good
Why call them Vector when they are just points ?
Points are representing data but the comprasion is made by vectors drawn between 2 data points. Comprassion and finding similarities are whole points of these databases.
That's the term used across many cs fields. Try a lil comp graphics and you'll find that any point in a n dimensional space is also called a vector