用自适应长上下文建模预测交通事故，精度达94%。
ALCo-FM: Adaptive Long-Context Foundation Model for Accident Prediction
- 根据数据波动性动态选取上下文窗口，提升长时序推理能力
- 在15个城市数据上实现0.94准确率、0.92 F1值，误差校准度仅0.04
- 适合城市交通风险预警与小样本迁移场景
交通事故虽发生率低但影响重大，需依赖长上下文多模态推理进行精准风险预测。本文提出ALCo-FM，一种统一的自适应长上下文基础模型，通过计算波动性预评分动态选择输入上下文窗口，并利用浅层交叉注意力编码融合多模态数据。在H3六边形网格上结合局部GAT层与BigBird风格稀疏全局Transformer，辅以蒙特卡洛丢弃实现置信度估计。模型在覆盖15个美国城市的训练数据上采用类别加权损失缓解标签不平衡问题，经少量数据微调后，在大规模城市风险预测任务中表现卓越，达到0.94准确率、0.92 F1值与0.04 ECE，超越20余种先进基线方法。代码与数据集已公开于：https://github.com/PinakiPrasad12/ALCo-FM。
原文摘要 · Abstract (English)
Traffic accidents are rare, yet high-impact events that require long-context multimodal reasoning for accurate risk forecasting. In this paper, we introduce ALCo-FM, a unified adaptive long-context foundation model that computes a volatility pre-score to dynamically select context windows for input data and encodes and fuses these multimodal data via shallow cross attention. Following a local GAT layer and a BigBird-style sparse global transformer over H3 hexagonal grids, coupled with Monte Carlo dropout for confidence, the model yields superior, well-calibrated predictions. Trained on data from 15 US cities with a class-weighted loss to counter label imbalance, and fine-tuned with minimal data on held-out cities, ALCo-FM achieves 0.94 accuracy, 0.92 F1, and an ECE of 0.04, outperforming more than 20 state-of-the-art baselines in large-scale urban risk prediction. Code and dataset are available at: https://github.com/PinakiPrasad12/ALCo-FM
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。