Primary competition visual

AirQo Ugandan Air Quality Forecast Challenge

Helping Uganda
$5 000 USD
Completed (almost 6 years ago)
Forecast
902 joined
327 active
Starti
Mar 14, 20
Closei
May 31, 20
Reveali
May 31, 20
Deep Learning starter notebook
Help · 12 May 2020, 22:36 · 14

Hi everyone,

I wrote a starter notebook for using Deep Learning on the competition data. You can run it on Google Colab in almost 5 mins training time and make a submission with ~41 RMSE on public leaderboard. I believe there are many ideas you can apply to have a better score.

I added also some data exploration to enable anyone who still didn't start to understand the data easily.

You can find the notebook here. Let me know if you have any idea to improve it or any question about the code.

Just remember to give me a star if you like it, please. :)

Cheers.

Discussion 14 answers

Thanks for sharing Karim

12 May 2020, 22:39
Upvotes 0

Thanks for sharing. Nice to see a different approach

13 May 2020, 05:03
Upvotes 0

Thanks. I hope it will be useful.

Thanks for sharing

13 May 2020, 08:34
Upvotes 0
User avatar
Mugisha_

@KarimAmer Just a slight correction.

It's not true that "Air quality causes thousands of deaths in Uganda", thats a very wrong allegation. You may have to edit your problem description. Otherwise great approach to problem solving.

13 May 2020, 17:15
Upvotes 0

Thanks steph for your comment. I used the numbers from State of Global Air website. I apologize for not adding the reference in the notebook.

However, it is not a pleasent introduction anyway so I changed it. Thanks for pointing this out to me.

Karim, thanks for the notebook!

13 May 2020, 19:29
Upvotes 0

You are welocme.

Thanks for sharing! I love your approach!

13 May 2020, 22:24
Upvotes 0

Thanks for your kind words.

I fixed some typos in data exploration figures. I hope it is more clear now.

14 May 2020, 14:47
Upvotes 0