From Scattered Samples to Fields of Distributions - Theory and Practice of Spatial Logistic Gaussian Processes

Explore the theoretical foundations and practical applications of Spatial Logistic Gaussian Processes in this seminar presented by Dr. Athénaïs Gautier from ONERA. Learn how to transform scattered data samples into comprehensive fields of distributions through advanced statistical modeling techniques. Discover the mathematical framework underlying spatial logistic Gaussian processes and understand their implementation in real-world scenarios. Examine the challenges of working with sparse or irregularly distributed data and master methods for creating continuous probability fields from discrete observations. Gain insights into the calibration and representation of prediction uncertainty in spatial modeling contexts. Understand the connections between classical statistical approaches and modern machine learning techniques for handling spatial data. This presentation is part of the "Representing, calibrating & leveraging prediction uncertainty from statistics to machine learning" research programme at the Isaac Newton Institute for Mathematical Sciences.