I was going through the "Forecasting At Scale" paper that explains Prophet (the time series forecasting tool developed by Meta). Whatever I could learn have tried to write down in a new blog post on my Substack. The goal was to break down each concept and understand the methodologies from scratch, along with the objective behind the approach taken: Please feel free to check it out:
https://t.co/mpzZC3Ubyi
(PS: Not sure what I have written...need to learn formatting properly and writing long posts without losing track ๐ฅฒ...feel free to point out typos, conceptual mistakes and whatever else you can find)
Columbia CS Prof explains why LLMs canโt generate new scientific ideas.
Bcz LLMs learn a structured โmapโ, Bayesian manifold, of known data and work well within it, but fail outside it.
But true discovery means creating new maps, which LLMs cannot do.
๐ ๐ ๐ฎ๐ฐ๐ต๐ถ๐ป๐ฒ ๐๐ฒ๐ฎ๐ฟ๐ป๐ถ๐ป๐ด ๐ก๐ผ๐๐ฒ๐ (๐๐ฎ๐ป๐ฑ๐๐ฟ๐ถ๐๐๐ฒ๐ป ๐ฃ๐๐)
Master the core of ๐ ๐ฎ๐ฐ๐ต๐ถ๐ป๐ฒ ๐๐ฒ๐ฎ๐ฟ๐ป๐ถ๐ป๐ด โalgorithms, models, training, evaluation & real-world examples. Perfect for interviews & AI enthusiasts! ๐ค๐
๐FREE for the first 500 people!
1. Like & Repost
2. Comment โMLโ
3. Follow (so I can DM you)
๐ก The future is machine learning, start learning today!
Also I am teaching a new course to nearly 200 graduate students this semester based on the new open-source textbook at the University of Hong Kong. We plan to post all the lecture slides, teaching materials, and video recordings on the book website too: https://t.co/leZlkURb7j Your contributions are welcome too!
@justinskycak Absolutely true. My coding skills became 10x more valuable when I combined them with domain expertise in my work. The intersection is where the magic happens.