Project suggestions for CS 4033/5033 Fall 2018

The following is a small list of suggestions. For other suggestions, visit previous year's project pages (linked from the main class page).

Ecology and sustainability projects

Imagine being in a forest with all the associated noises. Can you identify the bird calls? Can you identify what species of bird is making each call? Such data is available, talk to me if you are interested in this.
Vegetation or agricultural analysis: inagine a drone flying over a remote area. Can you identify the tree species? And the grass? Can you figure out the health of the farmland? This would be a nice vision/deep learning project.
Weather radar detects more than just weather. Detecting and identifying biological entities on the radar could be a fun project, though it would involve hand-labeling the data. If you are interested in the intersection of weather and ecology, this could be a fun project.

OU has purchases several minidrones and I have developed an open-source python library to control the drones. There are lots of interesting ML projects that can be done on the drones. Talk to me if you are interested in using one of the drones.

I have access to a tremedous amount of weather data. Let me know if you are interested in working on a weather-based project and I'll work to get you the data.

Kaggle has a variety of really need ML and data mining contests that are constantly running. It can be fun to pick one that is happening during the semester and participate in it. Who knows, maybe you can win big money while learning? No matter what, you can learn something from participating.
Note you cannot pick the handwriting recognition contest and data. This has been done in too many prior class projects and is no longer available as a project.

There are a lot of publicly available data sets ripe for learning from. For example, this is a meta-collection of data sets, meaning it is a collection of collections of data. You can find almost anything here!

Many of you are taking the class because you think machine learning can help you with your research. Feel free to use that research data as part of your class project. Just be sure you are willing to share the data with your partner as all the projects are partnered.

Spacewar is my implementation of the classical game of Asteroids. The code is very stable and lots of fun to play with. Feel free to use it as a simulation engine.

Last updated: August 5, 2018 6:34 PM