Title: Interpreting deep neural networks in a transformed domain
Abstract: Machine learning lies at the heart of new possibilities for scientific discovery, knowledge generation, and artificial intelligence. Its potential benefits to these fields require going beyond predictive accuracy and focusing on interpretability. In particular, many scientific problems require interpretations in domain-specific interpretable feature space (e.g. the frequency or wavelet domain) whereas attributions to the raw features (e.g. the pixel space) may be unintelligible or even misleading. To address this challenge, we propose TRIM (Transformation Importance), a novel approach which attributes importances to features in a transformed space and can be applied post-hoc to a fully trained model. We focus on a problem in cosmology, where it is crucial to interpret how a model trained on simulations predicts fundamental cosmological parameters. By using TRIM in interesting ways, we next introduce adaptive wavelet distillation (AWD), a method that aims to distill information from a trained neural network into a wavelet transform. Specifically, AWD penalizes feature attributions of a neural network in the wavelet domain to learn an effective multi-resolution wavelet transform. The resulting model is highly predictive, concise, computationally efficient, and has properties (such as a multi-scale structure) which make it easy to interpret. We showcase how AWD addresses challenges in two real-world settings: cosmological parameter inference and molecular-partner prediction. In both cases, AWD informs predictive features that are scientifically meaningful in the context of respective domains.
Short bio: Wooseok Ha is currently a postdoctoral researcher at the UC Berkeley Statistics Department under the supervision of Prof. Bin Yu. Previously, he was a Neyman Visiting Assistant Professor in the Department of Statistics at UC Berkeley, and a postdoctoral fellow at UC Berkeley's Foundation of Data Analysis (FODA) Institute and Berkeley Institute for Data Science (BIDS). Prior to that, he obtained his Ph.D. in statistics from the University of Chicago in 2018. His research interest includes high-dimensional statistics, optimization, interpretable machine learning, and graph clustering, as well as applications to other domains such as medical imaging, population genetics, and cosmology.