Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
WANGW09 3f4e65f5f4 | 2 years ago | |
---|---|---|
.META-INF | 2 years ago | |
data/METR-LA | 2 years ago | |
garage/metr | 2 years ago | |
images | 2 years ago | |
syn_batchnormal | 2 years ago | |
LICENSE | 2 years ago | |
TrafficPredictionExample.py | 2 years ago | |
config.yml | 2 years ago | |
engine.py | 2 years ago | |
generateSE.py | 2 years ago | |
generate_training_data.py | 2 years ago | |
model_dagstn.py | 2 years ago | |
model_dynamic_adj.py | 2 years ago | |
readme.md | 2 years ago | |
start_metr.sh | 2 years ago | |
test.py | 2 years ago | |
train.py | 2 years ago | |
util.py | 2 years ago |
动态注意力图时空网络(DAGSTNet)
交通流预测:根据过去一段时间的交通状况,预测未来一段时间的交通状况
以下三个数据集上的预测一个小时的平均性能指标:
数据集 | MAE | RMSE | MAPE |
---|---|---|---|
PeMS-BAY | 1.55 | 3.44 | 3.51% |
METR-LA | 3.00 | 5.99 | 8.12% |
PeMS-S | 2.57 | 5.07 | 6.42% |
MAE(Mean Absolute Error)
RMSE(Root Mean Squard Error)
MAPE(Mean Absolute Percentage Error)
PeMS-BAY comes from California Transportation Agen330 cies (CalTrans) Performance Measurement System (PeMS). This dataset contains 325 sensors in the Bay Area and records 6 months of data ranging from Jan 1st,2017 to May 31st, 2017.
METR-LA is collected from loop detectors on the high way of Los Angeles County. It contains 207 sensors and records 4 months of data ranging from Mar 1st, 2012 to Jun 30th, 2012.
PeMS-S is collected from the Caltrans Performance Measurement System (PeMS) and the data is collected from the California state highway system. The time range of PeMS-S is the weekdays of May and Jun of 2012 and 228 ensors are selected.
代码运行的环境与依赖。如下所示:
名称 | 版本 |
---|---|
pytorch | 1.4.0 |
numpy | 1.18.1 |
pandas | 1.1.3 |
matplotlib | 3.2.0 |
代码的输入与输出。如下所示:
数据文件夹中包含:train.npz,val.npz,test.npz 分别表示训练,验证,测试集。格式是[L,12,N,2].其中L表示样本数,12表示时间步个数,N表示路段个数,2表示通道维度(速度,时间)。由参数--data 指定文件夹位置。
邻接矩阵文件:adj_mx.pkl。表示路段之间的关系邻接矩阵文件。格式[N,N]。通过参数--adjdata指定文件位置。
节点嵌入文件:SE_file.txt。表示路段的嵌入。格式[N,64]。64表示一个路段使用64维的通道编码。由--SE_file指定文件位置。
模型的输出是[L,N,12] 表示预测的12个时间步的速度。保存在--test_output_file指定文件位置
在terminal下运行以下命令。
cd project_dir
python TrafficPredictionExample.py --main
python TrafficPredictionExample.py --test
generate_training_data.py
generateSE.py
动态注意力图时空网络(DAGSTNet)
Pickle Text Python other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》