Distributed Learning, Recognition, and Prediction by ART and ARTMAP Neural Networks

Category regions as new geometrical concepts in Fuzzy-ART and Fuzzy-ARTMAP

“…Moreover, the same geometrical concepts can be utilized in the framework of virtually any other ART-based neural network architecture as an aid to understand these architectures and to derive theoretical results describing their behavior. Examples of such architectures include dART (Carpenter, 1997) and dARTMAP , Boosted-ARTMAP (Verzi et al, 1998), Micro-ARTMAP (Gomez first define the quantities…”

“…We only refer to a limited number of them: ARTEMAP (Carpenter & Ross, 1995), Gaussian ARTMAP (Williamson, 1996), dART (Carpenter, 1997), dARTMAP (Carpenter, Milenova, & Noeske, 1998), ARTMAP-IC (Carpenter & Markuzon, 1998), Boosted ARTMAP (Verzi, Heileman, Georgiopoulos, & Healy, 1998), Micro-ARTMAP (Gomez Sanchez, Dimitriadis, Cano Izquierdo, & Lopez Coronado, 2000), Topographic Attentive Mapping network (Williamson, 2001) and finally Ellipsoid-ART/ARTMAP (Anagnostopoulos & Georgiopoulos, 2001). The above contributions revolve around modifications and enhancements as well as around new approaches based on the concepts of the original FA and FAM architectures.…”

Dynamics of Projective Adaptive Resonance Theory Model: The Foundation of PART Algorithm

“…Consider the data set { (1, 2, 3, 4), (2,3,4,5), (3,4,5,6), (4,5,6,7), (6,7,8,9), (1,2,3,4), (3,4,5,6), (2,3,4,5), (6,4,5,6), (4, 2, 3, 1), (6, 7, 1, 2), (4, 5, 6, 7)} of 12 points in 4-dimensional space. Note that a few points appear twice.…”

“…ART1 self-organizes recognition categories for arbitrary sequences of binary input patterns, while ART2 does the same for either binary or anolog inputs. Some other classes of ART neural network architectures such as Fuzzy ART [12], ARTMAP [10], Fuzzy ARTMAP [13], Gaussian ARTMAP [26], and Distributed ART and Distributed ARTMAP [5], [6] were then developed with increasingly powerful learning and patten recognition capabilities in either an unsupervised or a supervised mode. Simply speaking, an ART network includes a choice process and a match process as its key parts.…”

ARTMAP-IC and medical diagnosis: Instance counting and inconsistent cases

“…When ART a makes a choice during testing (Q = 1 ), the ARTMAP-IC algorithm is equivalent to a fuzzy ARTMAP algorithm. However the original ARTMAP notation has been changed somewhat to clarify network functions and for consistency with a family of more general ART systems (Carpenter, 1996).…”