Internship: Interactive learning with explanations

Dr. Sivan Sabato, Department of Computing and Software, McMaster University

Discriminative Feature Feedback is a recent learning model that I studied in a series of recent works [1,2,3]. The idea in this model is that the learning algorithms obtains not only examples and labels as training data, but also explanations, in the form of features that explain why some pairs of examples are not labeled the same, although they might be considered similar. In this internship, the student will develop a new and efficient version of the algorithm for the stochastic setting [3].
I am looking for an intern who is interested in theoretical machine learning and who is excited about developing and studying new learning approaches. Strong background in probability and discrete mathematics is required. Funding may be available. A successful internship may lead to an offer for a PhD position.

References

[1] Dasgupta, S., Dey, A., Roberts, N., & Sabato, S. (2018). Learning from discriminative feature feedback. Advances in Neural Information Processing Systems, 31.

[2] Dasgupta, S. & Sabato, S. (2020). Robust Learning from Discriminative Feature Feedback. Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 108:973-982.

[3] Sabato, S. (2023). Improved Robust Algorithms for Learning with Discriminative Feature Feedback. Proceedings of The 26th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 206:1024-1036.