|
Stephen Bach
Assistant Professor
|
|
My latest research is on weakly supervised machine learning, in which the goal is to train models without hand labeled data. With the advent of data-hungry representation learning techniques like deep neural networks, curating labeled training data has replaced feature engineering as the most expensive and time consuming task in machine learning. Weak supervision aims to overcome this bottleneck. I also work on statistical relational learning and information extraction.
News
- Our work on new theoretical guarantees for adversarial multi class learning with weak supervision is accepted to ICML 2021! This builds on our AISTATS 2021 paper, which only handled binary classification, with a novel method and analysis.
- Our work providing theoretical guarantees for learning from weak supervision sources without any assumptions on their distribution is accepted to AISTATS 2021!
- I spoke at the AAAI workshop on Graphs and More Complex Structures for Learning and Reasoning about our work on zero-shot learning with common sense knowledge graphs.
- New pre-print available on few-shot learning with auxiliary data. We show how to adapt additional, semantically related data to better learn novel classes.
I lead the BATS machine learning research group. In the tradition of groups like LINQS and DAGS, BATS stands for "Bach's Awesome Team of Students."
- Reza Esfandiarpoor
- Nihal Nayak
- Jasper Solt (Co-advised with Jonathan Pober)
- Zheng Xin Yong
- Peilin Yu
- Ross Briden
- Chace Hayhurst
- George Hu
- Top Piriyakulkij
- Gaurav Sharma
- Tom Liu
- Jessica Dai (Undergrad, 2021, Arthur AI)
- Tiffany Ding (Undergrad + Master's, 2021, Ph.D. at UC Berkeley)
- Amy Pu (Undergrad, 2021, Google)
- Dylan Sam (Undergrad, 2021, Ph.D. at Carnegie Mellon)
- Berkan Hiziroglu (Master's, 2020, Amazon)
- Angie Kim (Undergrad, 2020, The New York Times)
- Esteban Safranchik (Undergrad, 2020, Ph.D. at U. Washington)
 |
Snorkel is a framework for creating noisy training labels for machine learning. It uses statistical methods to combine weak supervision sources like heuristic rules and task-related data sets, i.e., distant supervision, which are far less expensive to use than hand labeling data. With the resulting estimated labels, users can train many kinds of state-of-the-art models. Snorkel is used at numerous technology companies like Google, research labs, and agencies like the FDA. |
 |
Probabilistic soft logic is a formalism for building statistical models over relational data like knowledge bases and social networks. PSL programs define hinge-loss MRFs, a type of probabilistic graphical model that admits fast, convex optimization for MAP inference, which makes them very scalable. Researchers around the world have used PSL for bioinformatics, computational social science, natural language processing, information extraction, and computer vision. |
In spring semesters, I teach machine learning (CSCI 1420).
In fall semesters, I usually teach a seminar on learning with limited labeled data (CSCI 2952-C).