CEDAS Data Science Hackathon 2025

Who can participate

The CEDAS Data Science Hackathon 2025 is open for all Bachelor, Master, and PhD students at the UiB Department of Informatics, or / and students who are affiliated with CEDAS, or / and participants with an interest in expanding their experience within data science based on a practical problem. A limited number of participants can sign up – students who sign up first and those affiliated with CEDAS / UiB will be prioritized. 

Practical information

The competition starts on Friday, January 31, 2025 at 16:00 (room to be announced). The submissions of the participants’ contributions are due – at the latest – by Sunday, February 2, 2025, at 23:59. Attendance of the opening session on Friday is mandatory. After that, participants are free to decide if they would like to continue working at the venue or elsewhere.

Throughout the competition, it will be possible to ask questions to the organizing committee – either on-site or via Discord.

The winners will be announced at an award ceremony. All participants, the evaluation and organizing committees, as well as representatives from Eviny are invited to take part in the ceremony.

The competition

Participants will work on a dataset from Eviny representing charging sessions of electric vehicles. Participants will be challenged to explore, present and analyze the data, as well as do predictions on unseen data.

No previous knowledge in the energy domain is required but having some basic insights into methods from data science, machine learning, statistics, and/or visualization is beneficial.

Participants are encouraged to work in groups of up to four people – either formed up-front or on the first day of the event.

Award categories and evaluation

There will be two categories for the assessment: (1) Bachelor's level, and (2) Master’s + PhD level. Evaluation criteria are:

• for the exploratory analysis of the data set: level of general data science skills, quality of the code delivered, quality of statistics, quality of visualization
• for the predictive modelling: predictive performance on a holdout dataset with specified metrics will count towards the total score
• overall presentation of the results and creativity in proposed solutions

All contributions will be assessed by researchers working in related areas, and all participants will receive individual feedback.

Registration

If you are interested in joining the competition, then please register via this link: https://skjemaker.app.uib.no/view.php?id=17897898