Adaptive Resonance Theory

Fuzzy ART is a ANN architecture that can learn without forgetting. It is similar to human memory where people can recognize their parents even if they have not seen them in a while and have learned many new faces since. The theory was developed by Grossberg and Carpenter and includes various types such as ART 1, ART 2, ART 3, and Fuzzy ART. ART 1 is an architecture that can be used for clustering of binary inputs only. ART 2 improved upon the ART 1 architecture to support continuous inputs. Fuzzy ART, used in the present work, incorporates fuzzy set theory into the pattern recognition process.

The stand alone ART algorithm is available in Python at https://github.com/cbirkj/art-python.

art.match_choice(c, norm, normI, normT, m, chm, rho, beta)[source]

Checks match criterion Compute choice equation Discovers best choice

Parameters:
  • norm – minimum of input and templates
  • normI – norm of input
Returns:

returns category choice location
art.template_options_loop(cmax, chmax, ch, nc, m, chm, rho)[source]

Match Criterion

Parameters:
  • cmax – Maximum choice (initialized to be -1)
  • chmax – Match Criterion (initialized to be -1)
  • ch – Template choice
  • nc – Number of Categories
Return cmax:

Maximum choice template location
art.ART(I, T, m, chm, nc, min_calc, rho, beta, j)[source]

Train ART - Create Template Matrix

Parameters:
  • I – Input
  • T – Template
  • cmax – Maximum choice (initialized to be -1)
  • chmax – Match Criterion (initialized to be -1)