1 code implementation • ICML 2020 • Jize Zhang, Bhavya Kailkhura, T. Yong-Jin Han
This paper studies the problem of post-hoc calibration of machine learning classifiers.
no code implementations • 2 Dec 2020 • Jize Zhang, Bhavya Kailkhura, T. Yong-Jin Han
In this paper, we leverage predictive uncertainty of deep neural networks to answer challenging questions material scientists usually encounter in machine learning based materials applications workflows.
1 code implementation • 18 Jul 2020 • Ankur Mallick, Chaitanya Dwivedi, Bhavya Kailkhura, Gauri Joshi, T. Yong-Jin Han
In this work, we show that the uncertainty estimation capability of state-of-the-art BNNs and Deep Ensemble models degrades significantly when the amount of training data is small.
no code implementations • 16 Jul 2020 • Shusen Liu, Bhavya Kailkhura, Jize Zhang, Anna M. Hiszpanski, Emily Robertson, Donald Loveland, T. Yong-Jin Han
The scientific community has been increasingly interested in harnessing the power of deep learning to solve various domain challenges.
no code implementations • 30 Jun 2020 • Shusen Liu, Bhavya Kailkhura, Jize Zhang, Anna M. Hiszpanski, Emily Robertson, Donald Loveland, T. Yong-Jin Han
The scientific community has been increasingly interested in harnessing the power of deep learning to solve various domain challenges.
1 code implementation • 16 Mar 2020 • Jize Zhang, Bhavya Kailkhura, T. Yong-Jin Han
We show that none of the existing methods satisfy all three requirements, and demonstrate how Mix-n-Match calibration strategies (i. e., ensemble and composition) can help achieve remarkably better data-efficiency and expressive power while provably maintaining the classification accuracy of the original classifier.
1 code implementation • 16 Dec 2019 • Huichen Yang, Carlos A. Aguirre, Maria F. De La Torre, Derek Christensen, Luis Bobadilla, Emily Davich, Jordan Roth, Lei Luo, Yihong Theis, Alice Lam, T. Yong-Jin Han, David Buttler, William H. Hsu
This system meets computational information and knowledge management (CIKM) requirements of metadata-driven payload extraction, named entity extraction, and relationship extraction from text.
1 code implementation • 13 Oct 2019 • Ankur Mallick, Chaitanya Dwivedi, Bhavya Kailkhura, Gauri Joshi, T. Yong-Jin Han
Experiments on a variety of datasets show that our approach outperforms the state-of-the-art in GP kernel learning in both supervised and semi-supervised settings.
no code implementations • 5 Jan 2019 • Bhavya Kailkhura, Brian Gallagher, Sookyung Kim, Anna Hiszpanski, T. Yong-Jin Han
We also propose a transfer learning technique and show that the performance loss due to models' simplicity can be overcome by exploiting correlations among different material properties.