Traffic Prediction
Traffic prediction is a task involving the forecasting of traffic conditions, such as vehicle flow and travel time, in specific areas or on roads. As an important application of time series analysis, its core objective is to optimize transportation systems through historical data and real-time information, reduce traffic congestion, improve road usage efficiency, and enhance safety.
METR-LA
TITAN
PeMS07
STAEformer
PEMS-BAY
STD-MAE
PeMS08
PDFormer
PeMSD4
STD-MAE
PeMSD8
Hierarchical-Attention-LSTM (HierAttnLSTM)
PeMS04
PDFormer
EXPY-TKY
STD-MAE
PeMSD7
STG-NCDE
PeMSD7(M)
STD-MAE
LargeST
PatchSTG
NE-BJ
RGDAN
PeMSD3
PeMSD7(L)
STD-MAE
BJTaxi
ST-SSL
NYCTaxi
PeMS-M
SZ-Taxi
NYCBike1
NYCBike2
Beijing Traffic
MemDA
HZME(inflow)
HZME(outflow)
CorrSTN
PeMSD4 (10 days' training data, 15min)
DASTNet
PeMSD4 (10 days' training data, 30min)
PeMSD4 (10 days' training data, 60min)
PeMSD7 (10 days' training data, 15min)
PeMSD7 (10 days' training data, 30min)
PeMSD7 (10 days' training data, 60min)
PeMSD8 (10 days' training data, 15min)
PeMSD8 (10 days' training data, 30min)
PeMSD8 (10 days' training data, 60min)
Q-Traffic
hybrid Seq2Seq