Hi Fellow Blinders, I'm trying to make an ultimate thread of recommended ML/ML engineering resources. Contributions much appreciated. I'll add them to the OP. Looking for anything, from resources for learning machine learning theory to deep learning to the latest trends in ML engineering and systems. Thanks everyone! Blind Posts: List of papers recommended by a Pinterest employee: https://www.teamblind.com/article/ML-design-interview-3cYD0vdM Books: Convex Optimization (Boyd) Introduction to Information Retrieval (Manning) Elements of Statistical Learning Introduction to Statistical Learning Foundations of Machine Learning (Mohri) Machine Learning Yearning by Andrew Ng (https://d2wvfoqc9gyqzf.cloudfront.net/content/uploads/2018/09/Ng-MLY01-13.pdf ) Smola ( https://alex.smola.org/drafts/thebook.pdf ) Blogs: Airbnb ( https://medium.com/airbnb-engineering/tagged/machine-learning ) Amazon? Deepmind ( https://deepmind.com/blog ) Facebook ( https://ai.facebook.com/blog/ ) Google ( https://cloud.google.com/blog/products/ai-machine-learning, https://ai.googleblog.com/, https://www.blog.google/technology/ai/, Linkedin (https://engineering.linkedin.com/blog/topic/machine-learning) Netflix ( https://research.netflix.com/research-area/machine-learning ) Pinterest ( https://medium.com/pinterest-engineering/machine-learning/home, https://labs.pinterest.com/projects/machine-learning/ ) Quora (https://www.quora.com/q/quoraengineering/Machine-Learning-at-Quora) Stripe ? ( https://stripe.com/blog/engineering/ ), Twitter ( https://blog.twitter.com/engineering/en_us/topics/insights.html ) Uber ( https://eng.uber.com/research/?_sft_category=research-ai-ml ) Courses: Andrew Ng's original coursera course Andrew Ng's deep learning courses Bloomberg ML course Fast.ai https://developers.google.com/machine-learning/guides/rules-of-ml/ Misc: Kaggle.com Matrix and Gaussian Identity notes: https://cs.nyu.edu/~roweis/notes.html Papers: Software Engineering for ML (https://www.microsoft.com/en-us/research/uploads/prod/2019/03/amershi-icse-2019_Software_Engineering_for_Machine_Learning.pdf) Hidden Technical Debt in ML systems ( https://papers.nips.cc/paper/5656-hidden-technical-debt-in-machine-learning-systems.pdf ) KDD NIPS, I linked a single paper but it's a great source in general ( https://papers.nips.cc/paper/7595-probabilistic-matrix-factorization-for-automated-machine-learning.pdf ) People: Yoshua Bengio Fei-Fei Li Trevor Hastie Jiawei Han Surya Ganguli Michael Jordan Guy Lebanon Yann LeCun Geoffrey Hinton Gary Marcus Andrew Ng Robert Tibshirani Ashutosh Saxena Alex Smola Videos: Cracking the ML Design Interview: Sample questions: E-commerce, design a system that determines which site visitors should be sent a discount code. Design a system for ranking the search results (Linkedin jobs, amazon shopping, etc)
What’s annoying is getting an ML interview. Even if I go through this, I can’t get FAANG to give me an ML interview. I guess I need to do side projects?
But you’re at Google... I’m surprised to hear this. What about switching teams to get exposure to ML?
In the meantime is there anything you would add to the list? Especially ml systems design I feel like that area is harder to find resources for studying.
Coursera has a free course (50+ hrs of videos) on machine learning taught by a Stanford professor
can you please add the link of Cracking the ML Design Interview? which videos are these? thanks
I recently found out this post on #leetcode which seems comprehensive https://leetcode.com/discuss/interview-question/system-design/566057/machine-learning-system-design-a-framework-for-the-interview-day Also Chip pro’s github on ML system design in interesting
Great reads: https://github.com/chiphuyen/machine-learning-systems-design/blob/master/build/build1/consolidated.pdf https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/46178.pdf A practical course covering practical tips in high level: https://course.fullstackdeeplearning.com/ An interesting post with links to some company blogs: https://becominghuman.ai/machine-learning-system-design-f2f4018f2f8
thanks for sharing!
Thank you
"."
+1
Any resources for Machine Learning Design on Scale?? Say Job Recommendation from LinkedIn? Millions of Users and Millions of Jobs
Stay on top of NIPS papers. There was a pretty good paper on federated learning and compressed gradients not too long ago. Scaling recommendations has been well studied. Typically the feature vectors are sparse so the covariance matrix is sparse too. Start from there.
I would say read eng blogs and publications in industry track by these companies to see how they design such ML systems.
Great initiative OP. Blind is probably not the most collaborative tool for this, can you create/share a google doc?
I'd prefer people just post here, it will help with visibility when the content is directly on the thread. Not opposed to creating a google doc though, but let's see how it goes here first.
In addition to collaboration here can you please provide a read-only compilation in a Google doc or some other tool (Github repo would be best) it’s hard to follow it here. Thanks 🙏