Eric J Ma's Website

In this talk, inspired by my 2017 attempt at demystifying Bayesian deep learning, I now attempt to demystify graph deep learning.

Virtual fireside chat about how computational biology shows up in Bioengineering.

Me explaining exactly what I state in the title!👆

In which I talk about how software testing is awesome for open source and data science.

In this video, Hugo Bowne-Anderson and I chat about the revolutionary package manager Pixi. We discuss myexperiences with Pixi, emphasizing its advantages in package management, reproducibility, and collaboration within data science teams. Key topics include the benefits of lock files, simplifying Docker and GPU environments, and the impact of Pixi on reproducibility and developer efficiency.

In which I tell you why Bayesian deep learning is nothing really special, delivered at PyData NYC 2017.

How to do parameter estimation and case/control comparison in PyMC3, delivered at PyCon 2017.

My rant against canned statistical procedures, delivered at PyCon 2019.

In this talk, I share five key principles, learned from the software development world, that help accelerate data science project development and amplify impact.

The many places that graphs can be found, delivered at Big Data Boston 2016 and hosted by DataCamp.

How data scientists can incorporate testing into their development, delivered at PyData Ann Arbor in Jan 2020.