Machine Learning
1297 Threads
*Geometric mixtures* of two pdfs= Bhattacharyya arc=1D exponential family with sufficient statistic=log density ratio, log-normalizer minus ฮฑ-Bhattacharyya distance. #NeurIPS2020...
How to learn Machine Learning? ( as recommended by the TensorFlow team at Google ) ๐งต๐
Here is every course that I've taken over the last 5 years to work full-time in Machine Learning applications: ๐งต๐
Learn JavaScript ๐ฅ It's the safest bet you can do. You can build almost anything with it: ๐น Front End ๐น Back End ๐น Mobile Apps ๐น Desktop Apps ๐น Machine Learning ๐น DevOps Let's g...
Not sure I fully buy โour goal as machine learning engineers should be to raise, rather than beat, human level performanceโ โ I think this is a temporary solution to get immediate...
Here's what your first 30 days of Machine Learning should look like. (I wish I had this before) ๐งต๐ https://t.co/7LsR7VsKB8
as a practitioner who 1) threw a vanilla transformer at a large time series dataset 2) got โgood enoughโ results 3) didnโt take the time to deeply understand the architecture or f...
Unit testing for ML pipelines is challenging given changing data, features, models, etc. Changing I/O make it hard to have fixed unit tests. To hackily get around this, I liberal...
Starting out in Tech & thinking to keep busy? Here are tutorials - updated (with low data vids) or you can download the PDF versions for courses in: -Machine Learning -Java Techn...
The math for machine learning always scared me. Until... This year when I across these free resources which helped me in a massive way! Here's everything you need to know about...
Going to try another way of explaining why I think ML product dev is broken, this time with a clear software analogy: (1/6) https://t.co/4jxQ9Ny0yg
Imagine you are in a machine learning interview and want to share your experience building deep learning projects. Donโt try to impress your interviewer by telling them how you app...
Iโve noticed some companies offering โtieredโ versions of their ML API products. It makes sense to me from a training/inference cost perspective but no sense to me from a customer...
๐๐ต ๐ต๐ฐ๐ฐ๐ฌ ๐ฎ๐ฆ 2 ๐บ๐ฆ๐ข๐ณ๐ด ๐ต๐ฐ ๐จ๐ฆ๐ต ๐ช๐ฏ๐ต๐ฐ ๐ฎ๐ข๐ค๐ฉ๐ช๐ฏ๐ฆ ๐ญ๐ฆ๐ข๐ณ๐ฏ๐ช๐ฏ๐จ. Yes, 2 years. Here some takeaways from my machine learning journey which you can learn from. (so that you don't spend 2 years li...
People often ask me how to build better intuitions about different machine learning and deep learning methods. This is a thread about my experience (as an NLP Researcher) building...
@seanjtaylor I think about this two ways: * what are the eng tools we need to facilitate multiple people working on ML for the same prod solution? CI/CD, @MLflow model promotion,...
A few things that could potentially give you an edge when applying for machine learning related jobs:
I have been thinking a lot about designing systems for reproducibility in ML experiments from a lens of identifying the right pain points, realistic solutions, and good UX.
Before you jump into deep learning, I would strongly advise you to do a few introductory machine learning courses to get up to speed with fundamental concepts like clustering, regr...
practical MLE tip: if you know your distribution isnโt Gaussian, min-max normalize instead of standardize https://t.co/zP6ivOFy7d
In good software practices, you version code. Use Git. Track changes. Code in master is ground truth. In ML, code alone isn't ground truth. I can run the same SQL query today and...
i preach โiterate on the input data, not the modelโ a lot, but i want to add a recent reflection: *not all raw data needs to be featurized and fed to a model*
every morning i wake up with more and more conviction that applied machine learning is turning into enterprise saas. iโm not sure if this is what we want (1/9)
Some things about machine learning products just baffle me. For example, I'm curious why computer vision APIs release "confidence scores" with generated labels. What's the business...