Experimental Comparisons of Semi-Supervised
Michael Georgiopoulos, Georgios C. Anagnostopoulos, Madan Bharadwaj · 2003
We present a series of experimental results that reveal the merits of the semi-supervised learning when applied to two different types of ART architectures (Fuzzy ARTMAP and Ellipsoidal ARTMAP). The concept of semi-supervised learning (SSL) was first introduced in the Simplified Boosted ARTMAP architecture by Verzi, et al., 2002, and was extended to Boosted Ellipsoidal ARTMAP by Anag-nostopoulos, et al., 2002. Semi-supervised learning (SSL) refers to the semi-supervised manner, according to which exemplars are formed during training to identify clusters. According to the typical, fully supervised learning scheme of ART ar-chitectures, training patterns that are similar to an already