A Latent Space Metric for Enhancing Prediction Confidence in Earth Observation Data
Ioannis Pitsiorlas, Argyro Tsantalidou, George Arvanitakis, Marios Kountouris, Charalampos Kontoes · 2024
A new approach for estimating confidence in machine learning model predictions, specifically in regression tasks utilizing Earth observation data with a particular focus on mosquito abundance (MA) estimation, is proposed here. We leverage the Variational AutoEncoder architecture to derive a confidence metric by the latent space representations of Earth observation datasets. This methodology is pivotal in establishing a correlation between the Euclidean distance in latent representations and the absolute error in individual MA predictions. Our study focuses on Earth observation datasets from the Veneto region in Italy and the Upper Rhine Valley in Germany, considering areas significantly affected by mosquito populations. A key finding is a notable correlation of 0.46 between the absolute error of MA predictions and the proposed confidence metric. This correlation signifies a robust, new metric for quantifying the reliability and enhancing the trustworthiness of the AI/ML model predictions in the context of both Earth observation data analysis and mosquito abundance studies.