A sharing of my code on Tianchi Traffic-Prediction-Contest
Using Xgboost model to make the prediction, finally we got the rank 28/1716.
You can load the origin data on this URL:
https://tianchi.aliyun.com/competition/information.htm?spm=5176.100067.5678.2.32335fddSZSIW2&raceId=231598
Update on March 19th. 2021:
When I look through this code written down around 4 years ago, I am kind of impressed. This was the frist time I have tried to involve in a data mining competition, the code is messed up and the result is not proud but it works. The most interesting part is I just use the general Python code to do the complicated feature engineering part without Pandas and Numpy, but the speed and memory cost are both quite good! I must appoligize this code is definately hard to read. Hopefully you can find something helpful from this dirty shit to anything you are working on.
Best,
Chuanyu
It is my first time I have participated in bigdata contest, Im sorry I have writen this code ugly and hard to understand.
*If you have some questions or suggestions just send them to my E-mail: [email protected] and i will write back soon *