Figure/Table detail

ST-Crime: A Retrieval Augmented Pre-trained Foundation Model for Environment-Dependent Crime Spatio-Temporal Prediction
WANG Tao, ZHANG Yifan, CHEN Peng
Journal of Geo-information Science, 2026, 28(1): 209-221.   DOI: 10.12082/dqxxkx.2026.250448

城市 评估指标 LR LSTM MiST CF AGL-STAN UrbanGPT UniFlow ST-Crime
纽约 Macro-F1 0.629 9 0.667 0 0.675 2 0.683 3 0.728 3 0.688 5 0.691 2 0.739 7
Micro-F1 0.618 4 0.524 1 0.557 2 0.576 8 0.679 3 0.579 2 0.620 0 0.687 1
洛杉矶 Macro-F1 0.345 7 0.421 5 0.430 5 0.439 1 0.570 8 0.586 0 0.618 1 0.643 3
Micro-F1 0.312 4 0.349 4 0.347 1 0.370 3 0.541 9 0.524 1 0.567 4 0.601 8
旧金山 Macro-F1 0.142 9 0.521 7 0.532 0 0.548 1 0.586 3 0.556 7 0.624 9 0.665 2
Micro-F1 0.141 5 0.364 7 0.388 7 0.352 3 0.450 4 0.468 5 0.490 3 0.537 5
Tab. 4 Prediction performance with sufficient training samples
Other figure/table from this article