I've recently been involved in hiring data scientists and I wanted to share some of my own reflections on the process. In particular, the things that I think are within the realm of a candidate's control, and how candidates can stand out.


ericmjl boosted

The idea that new code is better than old is patently absurd. Old code has been used. It has been tested. Lots of bugs have been found, and they’ve been fixed. @spolsky@twitter.com


I'm just continually blow away by what we can do with a bit of creativity. JAX _really_ enables high performance NumPy computing.

I've been head deep into the world of autoregressive hidden Markov models at work. It's brought me all over the place to Dirichlet processes, mixture models, and more. The more I read, the more I know I don't know about this part of the modelling world. Multiple times my mental models have had to be revised. My work is fun, this is a rare privilege!

ericmjl boosted

RT @KiwiEV@twitter.com

Now, if this isn't the coolest thing you've seen today then you need to close the refrigerator. It's an animation of how bridges were built in Central Europe in the middle-ages.

ericmjl boosted

RT @Rainmaker1973@twitter.com

The potato paradox is a mathematical calculation that has a counter-intuitive result.

"You have 100kg of potatoes, which are 99% water by weight. You let them dehydrate until they're 98% water. How much do they weigh now?"

The surprising answer is 50kg


What do Fermi estimation and the principled construction of Bayesian priors have in common? I try to explore the idea in my latest blog post: ericmjl.github.io/blog/2020/10

ericmjl boosted

RT @peterseibel@twitter.com

How to be a 10x engineer: help ten other engineers be twice as good.

ericmjl boosted

RT @qri_io@twitter.com

Qri is ultimately about collaborating on datasets, and the command line is a great place to start as it shows off a lot of what qri can do:


ericmjl boosted

RT @eugeneyan@twitter.com

Unpopular view: Data scientists should be more end-to-end.

While this is frowned upon (too generalist!), I've seen it lead to more context, faster iteration, greater innovation—more value, faster.

More details and Stitch Fix & Netflix's experience 👇


ericmjl boosted

Simplicity is not simple. The Danger of "Simplicity", by @asthasr@twitter.com.


ericmjl boosted

In particular, the use of Bootstrap 4 cards is amazeballz!

Show thread

I redid my personal website theme, and I'm quite liking the results. ericmjl.github.io

It’s a new year, and I decided to refresh my personal website! Starting with a dark theme: ericmjl.github.io/

ericmjl boosted

RT @jthandy@twitter.com

🔥 Hot off the press: “🎉 🥳🥂 Data Science Roundup Top 20 Posts of 2019! [DSR #211]” roundup.fishtownanalytics.com/

ericmjl boosted

🌧️ 🌧️ 🌧️
🌲🌳🌳 🌳 🌳 🌳
🌳 🌳 🌳
🐌 🍄

Show older

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!