National Data Science Bowl: Predict Ocean Health

Enter the 1st ever National Data Science bowl, with 175K in prizes and build an algorithm to automate the plankton image identification across 100+ classes. Plankton are critically important to ecosystem, but traditional methods for measuring their populations are time consuming and cannot scale for large-scale studies.

National Data Science Bowl Booz Allen and Kaggle are giving data scientists the opportunity to use their power for global good with the launch of the first-ever

National Data Science Bowl (NDSB). The NDSB is an online competition that will challenge the data science community to develop an algorithm to advance the study of marine biology and ocean health. Participants will have 90 days to create an algorithm and the winners will get a share of the $175k prize – the largest ever purse on the community for a project focused on social good.

Traditionally, identifying marine life involved dragging a net through the ocean, and collecting (and killing) intact specimens so that they could be manually tagged. A new technology called the In Situ Ichthyoplankton Imaging System (ISIIS) instead acts as a Xerox system for the ocean by scanning organisms in their natural environment.

The only problem is, it takes so many pictures that reviewing them all would take years.

Through the NDSB, data scientists will be given 100,000 images of plankton from the Hatfield Marine Science Center and asked to develop an algorithm that will enable real time analysis and classification of the plankton within these images.  Think of it as facial recognition technology for plankton.

The algorithm will help researchers be able to make real-time assessments on ocean health or management, like being able to predict whether the fishing industry will have a good year or how a tropical storm impacted the marine ecosystem.