@jay_mlr

Jay. Machine Learning Researcher. Non-authoritative source of ML quips and the occasional useful tweet.

cs.umd.edu/~jay
Joined November 2013
  • Tweets19
  • Following128
  • Followers163
Love PAC
1
Punk ROC curve
BarRNN
What are you gating for?
It's all been done before - no parody required
Any weight you want it, that's the way we'll anneal it- @BenjoLondon
1
Convolve Together
We're coming up with Imposterior parody titles
1
1
“Try to find a universal definition for a forest. Tree-cky!” -- @pascalhitzler speaking at KIT
1
5
I wish every research paper had a blog post like this with clear & simple explanations: xitter.cf/berty38/status/5953838…
Bert Huang 🐥@berty385 May 2015
Upcoming AISTATS paper ift.tt/1GWd0LO

Upcoming AISTATS paper

In about a week, our paper Unifying Local Consistency and MAX SAT Relaxations for Scalable Inference with Rounding Guarantees by Steve, me, and Lise will appear at AISTATS in San Diego. (Steve will…

berthuang.wordpress.com
1
3
All exchangeable variables are alike; each conditionally independent variable is conditioned in its own way. Reading: arxiv.org/pdf/1405.0501.pdf
Interested in how Linked Data and Info Extraction can work together? @LD4IE2014 workshop posted their CFP oak.dcs.shef.ac.uk/ld4ie2014… #ISWC
Can MLNs learn how to make tea? Seems possible from today's reading group: ftp.cs.rochester.edu/u/kautz…
“You can order by truth value, which is a partial ordering, but it’s very shallow since the values are 0 and 1. It’s not a very tall order."
Wry: “I’ve been working in this field for a long time. I was working on Big Data before it was big - back when it was just data.” #s4bd13
Load more