Making Recommender Systems better.
Joined March 2011
- Tweets46
- Following1
- Followers38
- Likes34
Fixed our text following reviewers comments, final preprint: arxiv.org/pdf/2104.05047.pdf.
Interesting to note that if stability constraints are not imposed, #recsys inconsistencies increase as latent space grows, i.e., exactly when recommendations are expected to be more personalized
Nice to see that the book we contributed a chapter in is listed by @bookauthority as the recommended reading on the #recsys topic for this year! Thanks to @slavaxx, @icantador, and @domonkostikk for making this book! bookauthority.org/books/new-…
Will be virtually presenting our #recsys2020 work today at the poster session starting at midnight (Moscow time). It's gonna be a challenge, never had to do this so late 🌆 . Also hope my neighbours won't hear me 💤🌛
This year #RecSys2020 features teaser videos and links them to the accepted contributions! Have a look at our teaser at vimeo.com/456557591
Evgeny Frolov retweeted
Glad to announce our paper "Stable Low-rank Tensor Decomposition for Compression of Convolutional Neural Network" has been accepted to #ECCV2020! Please, join our poster zoom sessions today (August 25) at 6.00-8.00 am (UTC + 1) and 14.00 - 16.00 (UTC + 1) arxiv.org/abs/2008.05441
Evgeny Frolov retweeted
Excellent work by Evgeny Frolov, Leyla Mirvakhabova & Valentin Khrulkov.
Better geometry beats model complexity. That's what we've discovered in our recent work "Performance of Hyperbolic Geometry Models on Top-N Recommendation Tasks" accepted at #recsys2020. More details in our pre-print: arxiv.org/abs/2008.06716, source code: github.com/evfro/HyperbolicR….
Better geometry beats model complexity. That's what we've discovered in our recent work "Performance of Hyperbolic Geometry Models on Top-N Recommendation Tasks" accepted at #recsys2020. More details in our pre-print: arxiv.org/abs/2008.06716, source code: github.com/evfro/HyperbolicR….
It's nice to find a concise description of our HybridSVD work among #recsys2019 highlights by Criteo team. Thanks for mentioning it, @CriteoAILab!
Understanding reliability of evaluation metrics is an important direction for research. Hope to see more work in this direction in the future.
Is your metric good for tuning your Recommender System? Look at arxiv.org/pdf/1909.02523.pdf #recsys2019 @ACMRecSys @sisinflab @PolibaOfficial @SigOpt
Great intro by @crmpla67 at #recsys2019 ORSUM workshop: you have to be pragmatic.
Evgeny Frolov retweeted
Questioning the methodology is part of research. #recsys2019 It's good to see this kind of paper, even better when awarded a prize.
I've decided to support my talk at #recsys2019 with some extra tutorial (like 8 pages + poster + talk is not enough :)). It extends the main repo, linked to our #recsys paper, and considers another popular hybrid MF model not included in the paper eigentheories.com/blog/light…
Have you ever wondered why there's no hybrid variant of the famous PureSVD approach? Turns out it's very easy to make! Will be presenting our #recsys paper on HybridSVD accepted at #recsys2019 as a long paper. Browser-runnable code for reproducing results: github.com/Evfro/recsys19_hy…
Slides from my #recsys2016 talk "Fifty Shades of Ratings..." are now also online slideshare.net/EvgenyFrolov7…
Evgeny Frolov retweeted
Flavian Vasile: improve cold start next song prediction by adding the likelihood that artists are listened in sequence #recsys2016
