Skip to content

Linear regression machine learning and data science problem to predict bike number for the 2017 AIB DataHack - Conor Gildea and Niall Hunt

License

Notifications You must be signed in to change notification settings

NiallEHunt/AIB-DataHack-2017---TheHuntForGild

Repository files navigation

AIB DataHack 2017 - The Hunt For Gild

The Hunt for Gild take on the AIB Datathon 2017!

200 PhD, Masters and Undergraduate students competing on the day.

AIB DataHack Logo

Task

Predicting numbers of bikes (in a bike sharing scheme) used on a day with machine learning.

Predictions are based on knowing factors for each day such as temperature, wind, precipitation, is it a holiday...etc and applying our trained machine learning model, based on these factors, to predict the number of bikes.

Given

We were giving two years of previous data relating to the bike sharing scheme, to train and test our machine learning model. Some of this data was purposely unclean, so we had to adjust this before training our model on it.

Result

We are very proud of our resulting model, especially since we weren't as used to Python and Machine Learning, as other topics.

A lower score is better.

The top two teams got a score around 950.

Third place got a score of 1100.

We got a score of 1250.

👨‍💻 👨‍💻

👌Super proud 👌

The Hunt for Gild

Our team consists of Conor Gildea and Niall Hunt. Two second year computer science students studying in Trinity College Dublin at the time of this competition.

Conor Gildea: https://www.linkedin.com/in/csigildea/

Niall Hunt: https://www.linkedin.com/in/niallehunt/

About

Linear regression machine learning and data science problem to predict bike number for the 2017 AIB DataHack - Conor Gildea and Niall Hunt

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages