Jaan Altosaar

☰

  1. A recommendation modeling framework applied to food recommendation and recommending arXiv papers.

    Press: Columbia University Postdoc Society Blog

    J. Altosaar, R. Ranganath, and W. Tansey. RankFromSets: Scalable Set Recommendation with Optimal Recall. Stat, 2021.

      URL   PDF   Code    

  2. Recommendation system for longform journalism deployed on the AWS lambda free tier.

    R. Bansal, J. Olmstead, U. Bram, R. Cottrell, G. Reder, J. Altosaar. Recommending Interesting Writing using a Controllable, Explanation-Aware Visual Interface. Workshop on Interfaces and Human Decision Making for Recommender Systems, ACM Conference on Recommender Systems, 2020.

    URL   PDF   Code     Talk   Demo  

  3. A framework to compare pre-trained representations for downstream tasks.

    W. F. Whitney, M. J Song, D. Brandfonbrener, J. Altosaar, K. Cho. Evaluating representations by the complexity of learning low-loss predictors. arXiv, 2020.

    arXiv   PDF   Code    

  4. It is time to go to the hooding

    Altosaar, J. Probabilistic Modeling of Structure in Science: Statistical Physics to Recommender Systems. Ph.D. Thesis, Princeton University, 2020.

    URL   arXiv   PDF   Code     Slides  

  5. Using language models to predict hospital readmission

    Press coverage: VentureBeat

    K. Huang, J. Altosaar, R. Ranganath. ClinicalBERT: Modeling Clinical Notes and Predicting Hospital Readmission. ACM Conference on Health, Inference, and Learning, 2020.

    arXiv   PDF    

  6. You dropout, we noisin'

    A. Dieng, R. Ranganath, J. Altosaar, D. Blei. Noisin: Unbiased Regularization for Recurrent Neural Networks. ICML, 2018.

    URL   arXiv   PDF     Slides  

  7. Preventing poor local optima in variational inference

    Invited talks at Imperial College London, NIPS Approximate Inference Workshop Spotlight, Machine Intelligence Research Institute, and Bloomberg Machine Learning Group.

    J. Altosaar, R. Ranganath, D. Blei. Proximity Variational Inference. AISTATS, 2018.

    URL   arXiv   PDF   Code     Slides  

  8. New divergences for variational inference

    R. Ranganath, J. Altosaar, D. Tran, D. Blei. Operator Variational Inference. NIPS, 2016.

    URL   arXiv   PDF     Slides  

  9. Word embeddings for recommender systems

    Dawen Liang, Jaan Altosaar, Laurent Charlin, and David Blei. Factorization meets the item embedding. Recommender Systems, 2016.

    URL   PDF   Code     Talk  

  10. Word embeddings applied to classical music

    Eamonn Bell, Jaan Altosaar. Word embedding models applied to classical music recover the circle of fifths in embedding space. ICML Music Discovery Workshop, 2016.

    URL   PDF   Code     Slides  

  11. We showed a pyrochloric oxide, considered to be a classical system, has quantum behavior.

    Featured on the PRB front page.

    P. Henelius, T. Lin, M. Enjalran, Z. Hao, J. Altosaar, P. Henelius, F. Flicker, T. Yavors’kii, and M. J. P. Gingras. Refrustration and Competing Orders in a Spin Ice Material. Phys. Rev. B., 2016.

    URL   arXiv   PDF   Code    

  12. Goal: model how words are used differently in arts versus sciences

    Jingwei Zhang, Aaron Gerow, Jaan Altosaar, James Evans, Richard Jean So. Fast, Flexible Models for Discovering Topic Correlation across Weakly-Related Collections. EMNLP, 2015.

    URL   arXiv   PDF   Code    

  13. MusicMappr: let anyone make a beat in 30 seconds in their browser

    Interview and live demo featured on The Wire magazine. Check out the live demo here!

    Ethan Benjamin, Jaan Altosaar. MusicMapper: Interactive 2D representations of music samples for in-browser remixing and exploration. New Interfaces For Musical Expression, 2015.

    URL   PDF   Code     Demo  

  14. Can we convey how fish move using generative music and computer vision?

    Andrew Mercer-Taylor, Jaan Altosaar. Sonification of Fish Movement Using Pitch Mesh Pairs. New Interfaces For Musical Expression, 2015.

    URL   PDF   Code     Demo