Fusion adaptive resonance theory (fusion ART)[1][2] is a generalization of self-organizing neural networks known as the original Adaptive Resonance Theory[3] models for learning recognition categories across multiple pattern channels.
There is a separate stream of work on fusion ARTMAP,[4][5] that extends fuzzy ARTMAP consisting of two fuzzy ART modules connected by an inter-ART map field to an extended architecture consisting of multiple ART modules.
In addition, various extensions have been developed for domain knowledge integration,[6] memory representation,[7][8] and modelling of high level cognition.
Fusion ART is a natural extension of the original adaptive resonance theory (ART)[3][9] models developed by Stephen Grossberg and Gail A. Carpenter from a single pattern field to multiple pattern channels.
Whereas the original ART models perform unsupervised learning of recognition nodes in response to incoming input patterns, fusion ART learns multi-channel mappings simultaneously across multi-modal pattern channels in an online and incremental manner.
Fusion ART employs a multi-channel architecture (as shown below), comprising a category field
The model unifies a number of network designs, most notably Adaptive Resonance Theory (ART), Adaptive Resonance Associative Map (ARAM)[10] and Fusion Architecture for Learning and COgNition (FALCON),[11] developed over the past decades for a wide range of functions and applications.
We show how fusion ART can be used for a variety of traditionally distinct learning tasks in the subsequent sections.
Using a selected vigilance value ρ, an ART model learns a set of recognition nodes in response to an incoming stream of input patterns in a continuous manner.
Fuzzy ARAM, based on fuzzy ART operations, has been successfully applied to numerous machine learning tasks, including personal profiling,[12] document classification,[13] personalized content management,[14] and DNA gene expression analysis.
During learning, fusion ART formulates recognition categories of input patterns across multiple channels.
The knowledge that fusion ART discovers during learning, is compatible with symbolic rule-based representation.
category nodes are compatible with a class of IF-THEN rules that maps a set of input attributes (antecedents) in one pattern channel to a disjoint set of output attributes (consequents) in another channel.
Due to this compatibility, at any point of the incremental learning process, instructions in the form of IF-THEN rules can be readily translated into the recognition categories of a fusion ART system.
Augmenting a fusion ART network with domain knowledge through explicit instructions serves to improve learning efficiency and predictive accuracy.
[16] For direct knowledge insertion, the IF and THEN clauses of each instruction (rule) is translated into a pair of vectors A and B respectively.
The vector pairs derived are then used as training patterns for inserting into a fusion ART network.
An instance of fusion ART, known as FALCON (fusion architecture for learning and cognition), learns mappings simultaneously across multi-modal input patterns, involving states, actions, and rewards, in an online and incremental manner.
A class of FALCON networks, known as TD-FALCON,[11] incorporates Temporal Difference (TD) methods to estimate and learn value function Q(s,a), that indicates the goodness to take a certain action a in a given state s. The general sense-act-learn algorithm for TD-FALCON is summarized.
The new Q-value is then used as the teaching signal (represented as reward vector R) for FALCON to learn the association of the current state and the chosen action to the estimated value.