Bryan Yates

My deep learning course at the University of Geneva is available on-line. 1000+ slides, ~20h of screen-casts. Full of examples in PyTorch.

fleuret.org/dlc/

And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)

fleuret.org/lbdl/

And here's the basis trade

Definitely need more! Added you to the list

For people who want more markets and trading in their lives, put together a starter pack with traders, quants, journalists, and more 📈 #finsky

With doge market cap at 59B, looks more valuable than about 64% of companies in the S&P

Visual Taylor Series Visualizations of Taylor Series, based on diagrams from Amie Wilkinson

Visualize infinite series! Here's a page I made a few years ago, letting you play with convergence and divergence.
www.bewitched.com/demo/taylor/...

Alright nerds, one like, one C++ tip/fact

Nvidia to replace Intel in Dow Jones Industrial Average Intel will lose its spot in the Dow Jones Industrial Average after a 25-year run to Nvidia , S&P Dow Jones Indices said on Friday, the latest blow to the struggling chipmaker that was among the first two technology firms to be included in the blue-chip index.

Wow, after 25 year run

H100 GPU

Lots of talented people to follow if you’re looking to up your software dev feed

The Elements of Statistical Learning

Very nice, so approximate the true posterior with a variational distribution and can then use a faster optimization technique. Will give the reference paper a read

Any recommendations to improve scaling for GPs on large data sets? Say 2+ million samples, many features. The prediction confidence is very nice to have

I’ll def check out the cloud hosting, thanks for the heads up. Though for now likely sticking with self host, the cloud end can definitely streamline a lot of work but being able to customize own hardware has some advantages depending on the use case

tmux + vim is all you need

Do it 💯 lots of quant trading applications for GPs

Ah very nice, you’re already in prod with it. Yeah it looks to be able to handle scaling well so far

Deleting code is very satisfying

Nice! ML community growing on here 🔥

Solid lineup 👇

Been exploring timescaledb for high-frequency trading data, nanosecond resolution timestamps, very large datasets. How’ve you found the documentation / ease of use?

Bluesky >> LinkedIn

Who is doing fun work with LLMs in statistics??

Command line familiarity is important for quant trading as well. Cluster use is increasing across trading firms

I try to do it before work, find that I can get a lot done in the early mornings, less distractions

After it doesn’t work and Claude can’t help

Know a few people are looking for ideas on how transformers can be applied to time series forecasting. Recommend checking out this paper.

arxiv.org/pdf/1912.09363