In this article, we propose an unsupervised feature selection algorithm based on the recently developed shrinking and expansion algorithm (SEA). The SEA is already an established, efficient algorithm ...