Geophysics Hackathon 2015
The 3rd Geophysics Hackathon will be in New Orleans, USA, on 17 and 18 October 2015. You are welcome to join us!
Thank you sponsors and supporters!
Many thanks to Karl Schleicher, who kindly supported the event.
- File:EMC Corporation logo.svg
What is a hackathon?
Hackathon is an event where people come together to collaboratively build and launch mobile or web apps aimed at solving a particular problem. They usually work in small groups of up-to five people over a day or two, the aim is to come up with a prototype at the end of the hackathon.
Guide to organizing a successful hackathon
The basic idea of the hackathon is to build free geophysics mobile apps or web-based applications we can show to the rest of the world — "Here's what creative scientists can do in 2 days".
The format will be an optional contest between teams of 4. Competing teams are eligible for prizes and glory. Please use EventBrite to register and stay up to date with the event.
Individuals or teams are welcome to come on a non-competitive basis, to learn, explore, or just to have fun.
We will be at Propeller, 4035 Washington Ave, New Orleans.
Sat and Sun, 17 and 18 October 2015 (the weekend right before the SEG Annual Meeting).
The hackathon is free to attend. Thank you to our generous sponsors for making this possible.
How to attend
Please sign up at EventBrite. Thank you!
You don't have to be in the room to play! We will be in a Google Hangout all weekend — watch this space.
In the meantime, please sign up at EventBrite.
These are completely hypothetical project ideas. Please steal/adapt/change/remove or add to them. They are not prescriptive, just intended to get the creative juices flowing. If you already have an idea, yours is probably better. Probably definitely.
- Compute a seismic quality or 'interpretation reliability' indicator from a seismic line or volume.
- Find and connect literature about machine learning in geophysics using online bibliographic data.
- Uncertain or stochastic wavelet extraction
- Fuzzy well ties, or 2D forward models.
- Dynamic Bayesian decision tree or value of information widget with fuzzy inputs.
- Amusing geophysics paper generator.
This could be a bottleneck... any input appreciated!
- We will have a well dataset from Nova Scotia: logs and cuttings data (regression, classification)
- We have a small text dataset from the journal 'Geophysics': authors, titles, abstracts (text mining, recommendation, co-author networks)
- We have a small text dataset from the SEG Annual Meeting: authors, titles, short abstracts (text mining, recommendation)
- I have a new rock property dataset, but it's rather sparse; I'm not sure what can be done with it yet
- We have plenty of seismic, but not a lot of labeled data... I will try to make some before the event (classification, regression)
- We may have another well dataset, but it's mostly well logs, no labelled data
Some image data would be awesome, if anybody knows of any! Especially core images...
Reading and inspiration
Here are some things you might like to read, to understand a bit about the theme. Please add more!
- Andrew Curtis — a researcher on the topic at Edinburgh, with lots of writings and links
- A Practical Primer on Geostatistics — book from USGS
- Shooting into the dark — blog post by Matt
- Do you know what you think you know — article by Matt
- Introduction to Geostatistics and Uncertainty - blog post by Maitri
- It's your hackathon. You can work on anything you like, in any way you like. If you don't want to be part of the contest, that's 100% great. Please just come and hang out!
- Team size. Teams in the contest should have no more than 4 members.
- Fresh code. None of your code should exist before the hackathon, and the code must be developed on site. Teams may plan ahead of time, however.
- Frameworks. You can use any language, frameworks, platforms, or libraries you wish. You can build an iOS or Android mobile app or a web app, but not a desktop application.
- Publishing code. There will be extra credit for published, open source code — but this is at the team’s discretion. We’d love to see a new software company too!
- Schedule. We will start as soon after 9 am on Saturday as we can. We will finish all work at 4 pm sharp on Sunday, and commence with demos.
- Prize categories. There will be at least 3 categories, something but not exactly like this: Best overall, Best idea, and Best code. There will be prizes.
- Judging criteria. To compete, you must deliver a working app, and a 5 minute demo/presentation. You will be scored on originality, functionality, science, impact, and design.
- Small print. By taking part in the event, you agree to allow Agile Geoscience to include mention of your participation in the event, and to appear in photos and video of the event.
- Smaller print. Agile Geoscience makes no claim over intellectual property generated during the event. Your work belongs to you and your team.
- Smallest print. Agile Geoscience reserves the right to change the rules before the events starts, and to adapt to the situation on the weekend itself, and is the sole arbiter of rule interpretation should disputes arise. Which seems unlikely. Let's face it, none of this is serious in any way.
- David Holmes
- Graham Ganssle
- Jamie Alison
- Greg Partyka
- Tracy Stark