Berkeley Undergrads use Machine Learning to Understand Concerns of Ocean Pollution
Link to Published Medium Article: https://medium.com/@subhikshamani/berkeley-students-use-machine-learning-resolve-concerns-of-ocean-pollution-adf36e72d2a3
Tackling floating debris in the ocean is a challenging problem to resolve. By the time plastic ends up in the middle of the ocean, most of it has been disintegrated down to small, microplastic-sized particles and sits suspended below the water’s surface. Before we can handle removing it from the ocean, our first task is to locate it. Identifying exactly where plastic is located is the first step to solving a larger issue, which starts off as floating debris, and eventually makes it way up to food chain to haunt the actions of us humans.
As an initiative to identify debris in the ocean, a group of six UC Berkeley students have utilized machine learning algorithms to identify a specific region of interest as characterized by its longitude and latitude to classify whether there’s plastic in the area or not. Their work primarily focuses on the Great Pacific Garbage Patch, and with the integration of more data, their machine learning architecture can be built upon to offer a solution to the ocean’s plastic pollution problem. The results are shared in a website, to support future work on marine debris identification and reproduction of results.
https://github.com/smani2/oceans-trash