🤖 AI Summary
This study investigates whether high uncertainty coverage can outperform precise prediction in improving path decision utility. Leveraging trajectory data from Beijing and Chengdu, it proposes a decoupled evaluation framework that employs a freezing protocol to disentangle speed prediction errors from actual path losses, combined with offline surrogate task construction, minimum-boundary path selection, and joint calibration techniques for empirical analysis. The work reveals that joint upper-bound coverage is not a reliable proxy for downstream path utility. Experiments demonstrate that although enhancing joint coverage improves uncertainty quantification, it paradoxically increases late arrival rates by 0.16 to 0.92 percentage points and elevates average travel time. These findings provide critical counterintuitive evidence for research on prediction-decision consistency.
📝 Abstract
Whether more accurate traffic forecasts or higher uncertainty coverage improve route decisions is unclear. We evaluate this question with a frozen protocol that separates speed error, joint candidate path upper bound coverage, route selection, and realized loss. Using processed road speed data from Beijing and Chengdu, we construct offline proxy tasks with 150 origin destination pairs, three candidate paths, and 14 test days per city. We compare raw 90th percentile path time bounds with jointly calibrated upper bounds under minimum bound route choice. Joint coverage rises from 83.19% to 92.26% in Beijing M1, from 75.14% to 88.33% in Chengdu M1, and from 74.01% to 90.64% in Chengdu M2. Yet C2 increases lateness by 0.1633, 0.7848, and 0.9200 percentage points, respectively, and mean travel time by 0.588, 3.082, and 4.418 seconds. In a separate Chengdu predictor comparison, a 14.91% reduction in speed mean absolute error accompanies a 1.4571 percentage point reduction in lateness under C0. Joint coverage is therefore not a surrogate for downstream route utility in these frozen tasks the offline results do not establish online or causal benefits.