SC19 Proceedings

The International Conference for High Performance Computing, Networking, Storage, and Analysis

Scalable Reinforcement-Learning-Based Neural Architecture Search for Cancer Deep Learning Research

Authors: Prasanna Balaprakash (Argonne National Laboratory), Romain Egele (Argonne National Laboratory), Misha Salim (Argonne National Laboratory), Stefan Wild (Argonne National Laboratory), Venkatram Vishwanath (Argonne National Laboratory), Fangfang Xia (Argonne National Laboratory), Tom Brettin (Argonne National Laboratory), Rick Stevens (Argonne National Laboratory)

Abstract: Cancer is a complex disease. There is a growing need for the design and development of data-driven and, in particular, deep learning methods for various tasks such as cancer diagnosis, detection, prognosis, and prediction. For nonimage and nontext cancer data, designing high-performing deep learning models is a time-consuming, trial-and-error task that requires both cancer domain and deep learning expertise. We develop a reinforcement-learning-based neural architecture search to automate predictive model development for a class of representative cancer data. We develop custom building blocks that allow domain experts to incorporate the cancer-data-specific characteristics. We show that our approach discovers deep neural network architectures that have significantly fewer trainable parameters, shorter training time, and accuracy similar to or higher than those of manually designed architectures. We study and demonstrate the scalability of our approach on up to 1,024 Intel Knights Landing nodes of the Theta supercomputer at the Argonne Leadership Computing Facility.

Presentation: file

Back to Technical Papers Archive Listing