wingsweihua / colight

CoLight: Learning Network-level Cooperation for Traffic Signal Control
http://personal.psu.edu/hzw77/publications.html
160 stars 58 forks source link

A quick question about the datasets #7

Closed IvanLLL closed 4 years ago

IvanLLL commented 4 years ago

Hi,

I notice that there are three volumns of Hangzhou datasets as well as Jinan datasets in the data folder.

Hangzhou:

Jinan

It is not very clear which ones were being measured in the performance table (i.e., Table 2) of your paper.

table

Thanks a lot.

wingsweihua commented 4 years ago

We provide those datasets for the researchers' need for more datasets. Our paper uses the ones without the number suffix.