Traffic Prediction on METR-LA

Metric: 12 steps MAPE (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Hide extra data

Sort:

#	Model↕	12 steps MAPE▼	Extra Data	Paper	Date↕	Code
1	DCGCN	9.94	No	Dynamic Causal Graph Convolutional Network for T...	2023-06-12	Code
2	STD-MAE	9.59	Yes	Spatial-Temporal-Decoupled Masked Pre-training f...	2023-12-01	Code
3	STGM	9.39	No	-	-	Code
4	T-Graphormer	8.62	No	T-Graphormer: Using Transformers for Spatiotempo...	2025-01-22	Code
5	TITAN	8.43	No	A Time Series is Worth Five Experts: Heterogeneo...	2024-09-26	Code

#1DCGCNSOTA
9.94
12 steps MAPE· 2023-06-12
Dynamic Causal Graph Convolutional Network for Traffic Prediction Code
#2STD-MAE
9.59
12 steps MAPE· Extra Data· 2023-12-01
Spatial-Temporal-Decoupled Masked Pre-training for Spatiotemporal Forecasting Code
#3STGM
9.39
12 steps MAPE
No paperCode
#4T-Graphormer
8.62
12 steps MAPE· 2025-01-22
T-Graphormer: Using Transformers for Spatiotemporal Forecasting Code
#5TITAN
8.43
12 steps MAPE· 2024-09-26
A Time Series is Worth Five Experts: Heterogeneous Mixture of Experts for Traffic Flow Prediction Code