Determine the Number of Unknown Targets in Open World Based on Elbow Method

TLDR

Generalized evidence theory extends Dempster‑Shafer theory to fuse uncertain information in open‑world settings, yet determining the number of unknown targets remains unresolved. This study proposes an elbow‑method‑based approach to estimate the number of unknown targets in open‑world evidence fusion. The method first identifies the open world, applies K‑means clustering to group categories, uses the elbow method to determine the correct number of unknown targets, updates the discernment frame, and evaluates performance through experiments. Experimental results demonstrate that the proposed approach outperforms existing methods in open‑world scenarios.

Abstract

Generalized evidence theory as an extension of Dempster-Shafer evidence theory can deal with uncertain information fusion in the open world. However, one of the open issues is to detect the number of unknown targets. In this article, a new method based on the elbow method is proposed to solve this problem. After the identification of the open world, K-means clustering is used to cluster categories. Then, the elbow method is used to find a correct number of unknown targets. The frame of discernment is reupdated. To test effectiveness of the proposed method, several experiments are conducted. The results illustrate the advantage of the proposed method in the open world.