These are blogs, sites, and newsletters that my colleagues read and I browse through. We can keep up with the rapidly increasing Machine Learning knowledge-base. These are good summations that guide us to the more detailed papers we should read in our specializations.
I think many researchers use arxiv Sanity Preserver, but it is talked about rarely. I use arxiv Sanity Preserver solely to find articles of interest. I use Mendeley to store, read, and markup those PDF-formatted articles that I find interesting through arxiv Sanity Preserver.
Over the last eight years, a newer phenomenon has been the uprising of blog sites that are great supplements to arxiv. The following are my “goto” blog sites:
- Medium is a myriad of blogs divided by class or category on-line daily publications. It is one of our “goto” blog sites. Also, Medium publications such as towardsdatascience, theStartup, machine-learning, Artificial Intelligence, and programming explain and summarize techniques and papers and expose us to what is new or a leading trend.
- The Deeplearning.ai blog site is a site that lists all the courses produced by Deeplearning.ai, a growing list of tutorials, the “Pie & AI” signup, and “Pie & AI “ event descriptions and dates. I recommend that you go to the Deeplearning.ai site at least once a month to review the new material.
- fast.ai site references blog, book, package, community chat group, and three years of courses ranging from beginner to SOTA (state-of-the-art) on deep learning. The package and courses are based on Pytorch, except for the first course based on Tensorflow. I expect another course version next year and another release of the fast.ai package. Last year they took a stab at Tensorflow on Swift. It will not surprise me if the upcoming course introduces Julia.
- realpython has the best Python tutorials I have ever read on advanced Python fundamentals. People that are Python beginners, as well as multi-year experienced Python software engineers, can learn from the tutorials on this site.
- goggle.ai.hub has components (mostly docker images), documentation, tutorials, Jupyter notebooks, code, Tensorflow examples, Kuberflow pipeline example codes, and much more, all centered on Machine Learning. The goggle.ai.hub contents can be used on your local computers and the cloud. It is not specific to the Google Cloud Platform (GCP).
- Kaggle is the site for Machine Learning competitions, all artifacts associated with the kaggle competitions, and various real-world domain datasets in Machine Learning. Many people have used the fast.ai package to learn machine learning techniques and that place them in the top 10% of any Kaggle competition.
- PapersWithCode will not stop you from searching GitHub for Python packages but will help you get the source code associated with a published machine learning paper. More critical, this community effort has both TRENDING and SOTA tabs. PapersWithCode is a fantastic compliment to arxiv Sanity Preserver.
The mission of Papers with Code is to create a free and open resource with Machine Learning papers, code, and evaluation tables. — https://paperswithcode.com/about
Other great machine learning blogs are:
- and add your favorite enterprise blogs.
- TheSequence Scope is a free subscription while the Edge has a $50/year subscription fee. The logo and quote state their mission well.
TheSequence is an unusual way to learn and reinforce your knowledge about machine learning and artificial intelligence.
The Algorithm is a newsletter for people who are curious about the world of AI. I’m here to help you cut through the nonsense and jargon to figure out what truly matters and where all this is headed. You’ll hear from me every Friday with updates and thoughts on the latest AI news and research (as well as some added magic and memes). — Karen Hao, Senior Reporter
These are useful podcasts, especially if you are in a situation where you can not read.
I find it useful to print these out. I am old school as you get Python doc snippets with a keystroke in most of the Python IDEs (Interactive Code Environments — code editors).