/AAAI2020-RiskOracle

AAAIA2020-Accident Forecasting Experiments Setups

Primary LanguageJupyter Notebook

AAAI2020-RiskOracle

AAAI 2020-Accident Forecasting Experiments Setups

-For detailed Codes and PKDE-based data transformation, please refer to Codes for TKDE: https://github.com/zzyy0929/Codes-for-RiskSeq-TKDE, Thank you very much!-

Dataset Sources

NYC dataset

Accident Records: https://data.cityofnewyork.us/Public-Safety/NYPD-Motor-Vehicle-Collisions/h9gi-nx95

Taxi Trip Records https://www1.nyc.gov/site/tlc/about/tlc-trip-record-data.page

Speed Values: http://data.beta.nyc/dataset/nyc-real-time-traffic-speed-data-feed-archived

Weather Records: https://www.wunderground.com/history/daily/us/ny/central-park/KNYC

Demographic Data: https://data.cityofnewyork.us/City-Government/Demograp-hic-Statistics-By-Zip-Code/

Road Network: https://data.cityofnewyork.us/City-Government/NYC-Street-Centerline-CSCL-/

SIP dataset

Due to the privacy protocols between USTC and SIP traffic administration offices, the statistics of traffic flows and speed values cannot be open source.

Here, we list the source of accident/event records we use in our experiments:

https://weibo.com/kzzxzhdd?is_all=1&stat_date=201701&page=2#feedtop