February and March of 2020. Sharif University of Technology, Tehran, Iran.
1153 Contestants|
121 Teams
86 Universities
The 2020 Data Days Competition held at Sharif university of technology. The website link The link to the documentation The link to the dataset You are given a dataset of 9 csv files and you are expected to sort adIds that are most likely to be clicked on by a certain user. The problem and the datasets are similar to this kaggle competiton.
You can chceck the jupyter notebook and see the algorithms and preprocessing we had done.
Team Pandas
Milad Soltany --- Linkedin, Github , Email
Abbas Omidi --- Linkedin, Github, Email
Amirhossein Heydarian --- Linkedin, Github, Email
feel free to reach out to us.
We tried out many different models and algorithms.
What worked the best was random forest with 400 estimators and max_depth of 15, you can check the preproccesings and features we extracted.
The final notebook which consists of all the codes.
Best.
Team Pandas.
About a year after the first match in which we succeeded to win second place in the country among more than 130 teams, we realized that we were, in fact, the first team in the country, but the refereeing team was realized this a little late and after announcing the final results.
Watch the relevant video below: