@backpropagating

PhD - HEAD OF DATA SCIENCE/ML/AI AND TECHNICAL LEAD / STRATEGIC VISION @ AgBiome - Machine Learning | Data Science | Biomedical Research

 North Carolina, USA
mutaku.io
Joined March 2011
  • Tweets2.5K
  • Following463
  • Followers89
33Media
Enter to Win Ruby Roubaix Gravel Fondo Event Packages #giveaway #win #gravel rubyroubaix.com/contest kingsumo.com/g/qmi3hf/enter-…
1
51
Goes for a lot of things in life.
Fine Woodworking@fwmagazine22 Nov 2022
On this, as so many other subjects, it might be a good idea to curb the obsessive need to tell others they are wrong. finewoodworking.com/2022/11/…

How many handplanes on the head of a pin? - FineWoodworking

On this, as so many other subjects, it might be a good idea to curb the obsessive need to tell others they are wrong.

finewoodworking.com
Matthew Martz retweeted
Julien Chaumond
@julien_c22 Nov 2022
is it me or has ML Twitter been way way quieter recently?
98
26
11
765
✨📔"We obtained 1.97M @GitHub repos with 9.06M @ProjectJupyter notebooks, 7.24M of which were unique. 69.5% of the 221M cells (~27B tokens) were code cells, while the remaining 30.5% were markdown cells." Look at these lovely t-SNE notebook clusters: 📄arxiv.org/abs/2201.12901
1
3
33
We are adding a new cool decorator in the Memray pytest plugin that will make your test fail if it leaks memory 🤖 Is like valgrind but a LOT of times faster 🚀and it doesn't get confused by the Python interpreter specialised allocator. And configure it is just one line! 🙀
10
85
8
504
Matthew Martz retweeted
Ben Hamner
@benhamner15 Nov 2022
Python dependency management
2
8
🤯🤯🤯🤯🤯🤯
Sebastian Raschka
@rasbt11 Nov 2022
Native Jupyter notebook diffs on GitHub. Yes yes yes! 👏
Super excited to release my new book Modeling Mindsets tomorrow 🥳 This is my first book since I decided to become a full-time writer. And it's the book I wish I had read a few years ago to save myself a lot of time. 1/n
10
75
3
561
Matthew Martz retweeted
PyCon US@pycon18 Oct 2022
Big news!! PyCon US is celebrating 20 years, come join us for the celebration. The #PyConUS 2023 website is live and Call for Proposals are now open! Sponsorship details for PyCon US are also available. Read all the details here: bit.ly/3VB7U3M
1
74
10
183
Pandas DataFrame output is now available for all sklearn transformers (in dev)! scikit-learn.org/dev/auto_ex… This will make running pipelines on dataframes soo much easier, and provides better ways to track feature names! thanks to @thomasjpfan @glemaitre58 and Christian Lorentzen!

Introducing the set_output API

This example will demonstrate the set_output API to configure transformers to output pandas DataFrames. set_output can be configured per estimator by calling the set_output method or globally by se...

scikit-learn.org
13
139
27
603
Matthew Martz retweeted
Bojan Tunguz
@tunguz6 Oct 2022
Announcement: deepmind.com/blog/discoverin… Nature paper: nature.com/articles/s41586-0… GitHub repo: github.com/deepmind/alphaten… #DeepLearning #MachineLearning #ArtificialIntelligence #Mathematics #Research #ML #DL #AI 4/4
2
7
3
61
Matthew Martz retweeted
Zen Moments
@Zen_Moments4 Oct 2022
It always seems impossible until it's done. ~ Nelson Mandela
1
22
85
❤️❤️
You’re unable to view this Post because this account owner limits who can view their Posts. Learn more
Matthew Martz retweeted
vicki@vboykis26 Sep 2022
The more a textbook cover looks like it’s been generated by an unhinged AI model that has completely misunderstood your prompt of “friendly fun fractal pattern” the harder the math inside is gonna be.
1
4
1
46
Backprop
An image of small plants propagating in plastic containers with images of backs crappily pasted on top of the plants

ALT An image of small plants propagating in plastic containers with images of backs crappily pasted on top of the plants

1
1
20
Matthew Martz retweeted
UCI@UCI_cycling24 Sep 2022
An attack for the history books 🌈 Just look how @AvVleuten powered away #Wollongong2022
56
421
211
3,544
Matthew Martz retweeted
vicki@vboykis20 Sep 2022
I don’t know how else to describe it other than moving quickly from one small fire to the next by order of priority. A lot of stuff both professional and family wise falls through the cracks.
Jason Jacobs
@jjacobs2219 Sep 2022
I honestly don’t know how anyone functions professionally with school-aged kids.
4
7
68
Matthew Martz retweeted
vicki@vboykis22 Sep 2022
Learn the classics and you'll never feel like you're not keeping up because it all builds on itself (matrix math, feature generation/extraction, supervised/unsupervised/performant computing etc.). Going deep on the details of a specific algo will never not serve you well, imho.
Christoph Molnar 🦋 christophmolnar.bsky.social@ChristophMolnar21 Sep 2022
Remember SVMs, anyone? What you learn today may be outdated tomorrow. What's your strategy in dealing with the reckless pace of machine learning?
5
15
99
Matthew Martz retweeted
Peter Wang 🦋
@pwang22 Sep 2022
Move code to data. That is the way. Increase the context-richness of the code, and you increase insight. Increase the domain expertise of the human writing the code, and you increase business ability.
Chris@criccomini21 Sep 2022
I don't think ETL/ELT is really a thing any more... between outbox pattern, CDC, kSQL, materialize, dbt, CDW, etc etc... data is extracted, transformed, and loaded all over the place... Maybe it always has been. Were we just lying to ourselves all these years? ETLELTETLELT....
4
1
11
Correlation Can Ruin Interpretability on Mindful Modeler mindfulmodeler.substack.com/…
1
9
1
31
Load more