A Study on Graph-Structured Recurrent Neural Networks and Sparsification with Application to Epidemic Forecasting

Abstract

We study epidemic forecasting on real-world health data by a graph-structured recurrent neural network (GSRNN). We achieve state-of-the-art forecasting accuracy on the benchmark CDC dataset. To improve model efficiency, we sparsify the network weights via transformed-${\ell }_1$ penalty and maintain prediction accuracy at the same level with 70% of the network weights being zero.

References in corpus (5)

• On the difficulty of training Recurrent Neural Networks
• Accurate estimation of influenza epidemics using Google search data via ARGO
• Graph-Based Deep Modeling and Real Time Forecasting of Sparse Spatio-Temporal Data
• Laplacian Smoothing Gradient Descent
• Deep Learning for Real-Time Crime Forecasting and its Ternarization