Tonight #NeurIPS2022, poster #611 - come chat w @crwhite_ml & me about 🦖RecZilla🦖.
No individual recommender works well on all datasets/metrics. But, metafeatures from datasets can be used to predict algorithm+hyperparam performance with respect to a metric.
1/
Check out our papers at #NeurIPS2022!
(1/3)
On the Generalizability and Predictability of Recommender Systems
Wed 4-6pm, #611
📜: https://t.co/3JvKXQxTlB
🦖: https://t.co/pB0UXGoeAX
📹: https://t.co/E8BLORhHSR
w/ @DuncanMcelfresh Jonathan, @KhandagaleSujay, @johnpdickerson
📝: On the Generalizability and Predictability of Recommender Systems
👥: @DuncanMcelfresh, @KhandagaleSujay, Jonathan Valverde, @johnpdickerson, and @crwhite_ml
📆: NeurIPS, Main Conference
🔗: https://t.co/UQAOAEiVNw (Paper)
🔗: https://t.co/42oc5HoRhn (Code)
this is a fantastic opportunity, especially for CS & math folks who want to engage deeply with philosophy and the cognitive sciences. these folks are excellent collaborators and mentors. go apply!!
new postdoc position available on moral artificial intelligence at @CarnegieMellon and/or @DukeU, with Hoda Heidari, Jana Schaich Borg, Walter Sinnott-Armstrong, and myself -- reach out soon if you can't get application in in time / please share! @AIESConf https://t.co/DJ0TP6d0rM
For more background, there is troubling (https://t.co/Hn082lkMBJ) & worrying (https://t.co/WatdRuKPaW) evidence that SoTA rec-sys algs don’t generalize, and don’t beat common baselines. These papers are also worth a read. we build on their codebase in our work (5/4)
On the Generalizability and Predictability of Recommender Systems.
📄:https://t.co/Tj1y6m32oR
🦖:https://t.co/P6AXAZTtqU
🎥:https://t.co/17sTk6Yxra
Joint with Jonathan Valverde, @KhandagaleSujay, @johnpdickerson, @crwhite_ml (1/4)
We release RecZilla - given a new dataset, it uses meta-features to predict which rec-sys algorithm will work best (for any one of 315 performance metrics). See our codebase, pretrained RecZilla models, and *all* of our raw experimental results here: https://t.co/P6AXAZTtqU (4/4)
Lots of people have been asking for my advice on grad school lately, & I always say this:
1) pick a lab by PI 👩🏼🔬 > project 🦠
2) get a therapist 🛋
3) don’t compromise sleep, working out, water, or vegetable intake ✅
4) you’re more important than your work 🙏
Excited for this new adventure with @Ci2iVA and @StanfordHP! Bay Area folks, if you want to talk applied CS, algorithms in healthcare, AI and ML ethics, or anything else, hit me up ☕️☕️☕️
How can data impact decision-making in the public sector and how should it be used responsibly? @jessie_fin interviews Katie Bridges, a Senior Business Intelligence Analyst for the City of Boulder to understand her experiences using data within this space. https://t.co/B6EzEVpwVq
How can data impact decision-making in the public sector and how should it be used responsibly? @jessie_fin interviews Katie Bridges, a Senior Business Intelligence Analyst for the City of Boulder to understand her experiences using data within this space. https://t.co/B6EzEVpwVq
I'm partway through reading the latest @NIST report on managing bias in AI and already my neck is starting to ache from the amount of vigorous head nodding I've done. So far, it is really, really good��and covers so much more than bias mitigation. Read it!
https://t.co/3yKOqRiVxO
��🩸Join us for the next talk in our #AI4SocialGood seminar series! @DuncanMcelfresh will speak about Algorithms for blood donation and resource allocation in healthcare🩸✨
📢 @DuncanMcelfresh
🗓️ Wednesday, 3/16
⏰ 11:30 AM EST
👉Register at: https://t.co/t7thqxhlqZ
✨🩸Join us for the next talk in our #AI4SocialGood seminar series! @DuncanMcelfresh will speak about Algorithms for blood donation and resource allocation in healthcare🩸✨
📢 @DuncanMcelfresh
🗓️ Wednesday, 3/16
⏰ 11:30 AM EST
👉Register at: https://t.co/t7thqxhlqZ