Data to the Rescue - Predicting and Preventing Accidents at Sea
Explore a groundbreaking conference talk on predicting and preventing maritime accidents through advanced data science and machine learning techniques. Dive into the revolutionary approach of using ship behavior data, including location, speed, maps, and weather information, to forecast potential disasters at sea. Learn how the 350-year-old marine insurance industry is being transformed by modern algorithms and data analysis. Discover the process of gathering fragmented ship movement information and converting it into powerful machine learning models. Gain insights into the challenges of introducing cutting-edge technology to a traditional industry and the importance of model interpretability using SHAP. Presented by Dr. Yonit Hoffman from Windward, this talk demonstrates how data science can significantly reduce the human, financial, and environmental costs associated with maritime accidents.