Rutgers Machine Learning Group

   Sungjin Ahn
   Assistant Professor

  • In Spring 2020, I'll be teaching CS 536: Machine Learning
  • Our paper on Generative Hierarchical Models for Parts, Objects, and Scenes is now in arXiv.
  • Our paper on Scalable Object-Oriented Sequential Generative Models is now in arXiv.
  • We have 3 papers accepted in NeurIPS 2019 including one spotlight paper. Check out our publications.
  • Learning Single-View 3D Reconstruction with Adversarial Training is accepted to ICCV 2019 as an oral presentation paper.
  • Our paper on Sequential Neural Processes is in arXiv! [pdf][project]
  • I'm teaching Pattern Recognition (a.k.a. Machine Learning I) in Fall 2019

About Me
  • I'm an Assistant Professor at the Department of Computer Science at Rutgers University where I lead the Rutgers Machine Learning Group. My research focus, probabilistic agent learning, centers around (1) deep learning, (2) probabilistic inference, and (3) brain-inspired learning algorithms to make an AI agent that can learn the model of the complex and interactive world like humans. I received my Ph.D. at the University of California, Irvine on the study of scalable approximate Bayesian inference under the supervision of Prof. Max WellingI did my postdoc working on deep learning at MILA under Prof. Yoshua Bengio. Then, I joined Rutgers University in Fall 2018.

  • CBIM 9, 617 Bowser Rd., Piscataway, NJ, 08854
  • sjn.[my_last_name] at

Research Focuses

For probabilistic agent learningI'm currently focusing on the following problems
    • Probabilistic Generative Models
    • Representation Learning
    • Meta Learning
    • Unsupervised structured perception
    • Unsupervised reinforcement learning (exploration and unsupervised skill acquisition)
using the following methodologies
    • Deep Learning
    • Bayesian Learning
    • Reinforcement Learning
    • Brain-inspired Learning