RNN with Particle Flow for Probabilistic Spatio-temporal Forecasting
This work addresses the need for uncertainty quantification in spatio-temporal forecasting for applications like wireless, traffic, and financial networks, representing an incremental improvement over existing deep learning models that focus only on point forecasts.
The paper tackled the problem of probabilistic spatio-temporal forecasting by modeling time-series data as a random realization from a nonlinear state-space model and using particle flow for Bayesian inference of hidden states. The result showed that the approach provides better uncertainty characterization while maintaining comparable accuracy to state-of-the-art point forecasting methods on real-world datasets.
Spatio-temporal forecasting has numerous applications in analyzing wireless, traffic, and financial networks. Many classical statistical models often fall short in handling the complexity and high non-linearity present in time-series data. Recent advances in deep learning allow for better modelling of spatial and temporal dependencies. While most of these models focus on obtaining accurate point forecasts, they do not characterize the prediction uncertainty. In this work, we consider the time-series data as a random realization from a nonlinear state-space model and target Bayesian inference of the hidden states for probabilistic forecasting. We use particle flow as the tool for approximating the posterior distribution of the states, as it is shown to be highly effective in complex, high-dimensional settings. Thorough experimentation on several real world time-series datasets demonstrates that our approach provides better characterization of uncertainty while maintaining comparable accuracy to the state-of-the art point forecasting methods.