Preserving Dynamic Attention for Long-Term Spatial-Temporal Prediction

TL;DR

This paper introduces DSAN, a novel neural network with multi-space attention that improves long-term spatial-temporal predictions by effectively filtering noise and reducing error propagation in urban data mining applications.

Contribution

The paper proposes a Dynamic Switch-Attention Network with Multi-Space Attention to explicitly measure correlations and dynamically filter irrelevant information for better long-term predictions.

Findings

DSAN outperforms existing models in long-term spatial-temporal prediction tasks.

The multi-space attention mechanism effectively filters noise and reduces error propagation.

Extensive experiments validate DSAN's superior performance in urban data prediction scenarios.

Abstract

Effective long-term predictions have been increasingly demanded in urban-wise data mining systems. Many practical applications, such as accident prevention and resource pre-allocation, require an extended period for preparation. However, challenges come as long-term prediction is highly error-sensitive, which becomes more critical when predicting urban-wise phenomena with complicated and dynamic spatial-temporal correlation. Specifically, since the amount of valuable correlation is limited, enormous irrelevant features introduce noises that trigger increased prediction errors. Besides, after each time step, the errors can traverse through the correlations and reach the spatial-temporal positions in every future prediction, leading to significant error propagation. To address these issues, we propose a Dynamic Switch-Attention Network (DSAN) with a novel Multi-Space Attention (MSA) mechanism that measures the correlations between inputs and outputs explicitly. To filter out irrelevant noises and alleviate the error propagation, DSAN dynamically extracts valuable information by applying self-attention over the noisy input and bridges each output directly to the purified inputs via implementing a switch-attention mechanism. Through extensive experiments on two spatial-temporal prediction tasks, we demonstrate the superior advantage of DSAN in both short-term and long-term predictions.