Information Fusion-2-Text: Explainable Aggregation via Linguistic Protoforms

摘要

Graphical illustration of fusion-2-text for the case study explored herein involving object detection and land cover classification in remote sensing. First, multiple machine learning (ML) models are trained in a cross validation context. Next, fusion is used to combat the fact that no single ML architecture is best across all data and classes. However, what have we learned? Fusion-2-text is used to discover a succinct set of important summaries or anomalies for a user in or across models.