Machine Learning Blog | ML@CMU | Carnegie Mellon University


99 publications. 14 categories. 65 tags.



Please review the author guidelines below and fill out our short submission form.

The editorial team will follow up with you shortly. Any questions can be directed to the editorial team at:

Author Guidelines


The purpose of this blog is to present machine learning research at CMU. The majority of posts will be on original research conducted by members of the machine learning community at CMU. Tutorials and surveys on a particular topic are also welcome, but we expect these to constitute a smaller portion of contributed posts.

Authors should keep in mind that the target audience is students / researchers / engineers with basic machine learning background. That is to say, authors can assume the readers know what is Naive Bayes, gradient descent, etc; however, for specific sub-area, e.g., nonparametric regression, authors will need to provide necessary background and motivation.


Since the goal of this blog is to present original research, we encourage authors to present their own papers in a more casual blog post format.  Anything related to machine learning / data science / AI is welcome.


If you would like to contribute a blog post, please contact us with a short description of the content (e.g., an abstract of your work). We will reply with next steps if we decide to proceed with your proposed blog post.

A primary editor will be assigned to you to help answer any questions you may have as you prepare your submission.  After you have a first draft, the primary editor will collect feedback from a few reviewers for you to incorporate in the final version.  Once your blog post is ready, it will go live on the website. Feel free at this point to promote your blog post through social media. We will also help promote posts from time to time through our own social network channels (Twitter / FB / wechat / etc.).


Every blog post should consist of a title and a main text.  In general, blog posts should have 1000 – 3000 words, which takes about 8 – 25 minutes to read.  

There is no strict format for the main text. We encourage authors to start with a high level description of the problem, followed by the author’s proposed solution and results, which will be dissected in the remainder of the post.  For longer submissions, a separate abstract is highly recommended, along with clearly delineated subsections.

It is crucial to properly motivate the problem by providing adequate context and citations for relevant related work.  Feel free to use hyperlinks throughout the blog post to link to useful sources.

Aside from text, we believe visual aids, when used appropriately, are highly effective in conveying information. We encourage authors to include thoughtful data visualization, charts, other graphics and even animations when suitable.  Authors are free to express their creativity through the added flexibility afforded by a blog in contrast to a static medium like paper.


Since this is a blog and not an academic paper, authors do not need to use a rigorous style of writing. In contrast to academic papers, we encourage authors to include more high level intuitions of problems, concepts, algorithms, equations, etc. Plots and animations are highly encouraged. See example blogs.

For the core concepts, authors can provide mathematical definitions but authors should focus on providing intuition for technical results as opposed to relying on equations to get the key ideas across.  Again, our target audience are those well versed in basic ML concepts, but not necessarily experts in ML/AI.

Since this is an academic blog hosted by CMU, authors should be careful to be as objective as possible when discussing his/her own and others’ research.  Please note that we will be appending the following disclaimer to the end of all posts:

“All opinions expressed in this posts are those of the author and do not represent the views of CMU.”