|
Abstract:
|
The frequency, intensity, and pervasiveness of climate-related natural disasters appear to be increasing with each passing year. Extreme heat, drought, wind, and wildfires can lead to food insecurity, socioeconomic instabilities, ecological disruptions, as well as property loss and casualties. By combining the efforts of physical scientists, data scientists, and software engineers, Jupiter Intelligence aims to provide high-resolution, forward-looking risk analysis of these climate perils and more. This talk will outline various facets of data science as they apply to our product development: modeling, data engineering, and system design and optimization. We will focus on scaling and deploying cloud-based computational pipelines as well as some of the data engineering challenges. The implementation of machine learning and statistical methods will be discussed in each component.
|
