Type Hierarchy and State
Internal storage conventions and model types. Public base types are documented under Advanced configuration.
Index
SingleART,AbstractFuzzyART,MergeART,opts_MergeARTARTMatrix,ARTVector,ARTStats,ARTIterator,ART_DIM,ART_SAMPLES
Model Types
SingleART
AdaptiveResonance.SingleART — Type
abstract type SingleART <: ARTSummary
Abstract supertype of single ART modules with matrix weights and symbol-based activation, match, and update functions.
Fields
AbstractFuzzyART
AdaptiveResonance.AbstractFuzzyART — Type
abstract type AbstractFuzzyART <: AdaptiveResonance.SingleARTSummary
Abstract supertype of FuzzyART modules.
Fields
MergeART
AdaptiveResonance.MergeART — Type
mutable struct MergeART <: ARTSummary
MergeART module struct.
For module options, see AdaptiveResonance.opts_MergeART.
References
- L. E. Brito da Silva, I. Elnabarawy, and D. C. Wunsch, 'Distributed dual vigilance fuzzy adaptive resonance theory learns online, retrieves arbitrarily-shaped clusters, and mitigates order dependence,' Neural Networks, vol. 121, pp. 208-228, 2020, doi: 10.1016/j.neunet.2019.08.033.
- G. Carpenter, S. Grossberg, and D. Rosen, 'Fuzzy ART: Fast stable learning and categorization of analog patterns by an adaptive resonance system,' Neural Networks, vol. 4, no. 6, pp. 759-771, 1991.
Fields
opts::opts_DDVFA: DDVFA options struct.
subopts::opts_FuzzyART: FuzzyART options struct used for all F2 nodes.
config::DataConfig: Data configuration struct.
threshold::Float64: Operating module threshold value, a function of the vigilance parameter.
F2::Vector{FuzzyART}: List of F2 nodes (themselves FuzzyART modules).
labels::Vector{Int64}: Incremental list of labels corresponding to each F2 node, self-prescribed or supervised.
n_categories::Int64: Number of total categories.
epoch::Int64: Current training epoch.
T::Vector: DDVFA activation values.
M::Vector: DDVFA match values.
stats::Dict{String, Any}: Runtime statistics for the module, implemented as a dictionary containing entries at the end of each training iteration. These entries include the best-matching unit index and the activation and match values of the winning node.
opts_MergeART
AdaptiveResonance.opts_MergeART — Type
mutable struct opts_MergeART <: ARTOptsSummary
MergeART options struct.
These options are a Parameters.jl struct, taking custom options keyword arguments. Each field has a default value listed below.
Fields
match_tracking::Bool: Flag to enable match tracking. Default: falseepsilon::Float64: Positive match-tracking increment: episilon ∈ (0, 1) Default: 0.001rho_lb::Float64: Lower-bound vigilance parameter: rho_lb ∈ [0, 1]. Default: 0.7rho_ub::Float64: Upper bound vigilance parameter: rho_ub ∈ [0, 1]. Default: 0.85alpha::Float64: Choice parameter: alpha > 0. Default: 0.001beta::Float64: Learning parameter: beta ∈ (0, 1]. Default: 1.0gamma::Float64: Pseudo kernel width: gamma >= 1. Default: 3.0gamma_ref::Float64: Reference gamma for normalization: 0 <= gamma_ref < gamma. Default: 1.0similarity::Symbol: Similarity method (activation and match): similarity ∈ [:single, :average, :complete, :median, :weighted, :centroid]. Default: :singlemax_epoch::Int64: Maximum number of epochs during training: max_epochs ∈ (1, Inf). Default: 1display::Bool: Display flag for progress bars. Default: falsegamma_normalization::Bool: Flag to normalize the threshold by the feature dimension. Default: trueuncommitted::Bool: Flag to use an uncommitted node when learning.If true, new weights are created with ones(dim) and learn on the complement-coded sample. If false, fast-committing is used where the new weight is simply the complement-coded sample. Default: false
activation::Symbol: Selected activation function. Default: :gamma_activationmatch::Symbol: Selected match function. Default: :gamma_matchupdate::Symbol: Selected weight update function. Default: :basic_updatesort::Bool: Flag to sort the F2 nodes by activation before the match phaseWhen true, the F2 nodes are sorted by activation before match. When false, an iterative argmax and inhibition procedure is used to find the best-matching unit. Default: false
Storage and Dimensions
ARTMatrix
AdaptiveResonance.ARTMatrix — Type
ARTMatrix
Description
The type of matrix used by the AdaptiveResonance.jl package, used to configure matrix growth behavior.
ARTVector
AdaptiveResonance.ARTVector — Type
ARTVector
Description
The type of vector used by the AdaptiveResonance.jl package, used to configure vector growth behvior.
ARTStats
AdaptiveResonance.ARTStats — Type
ARTStats
Description
Definition of the ART module statistics dictionary, used to generate and store various logs during training and testing.
ARTIterator
AdaptiveResonance.ARTIterator — Type
ARTIterator
Description
Acceptable iterators for ART module training and inference
ART_DIM
AdaptiveResonance.ART_DIM — Constant
ART_DIM
Description
AdaptiveResonance.jl convention for which 2-D dimension contains the feature dimension.
ART_SAMPLES
AdaptiveResonance.ART_SAMPLES — Constant
ART_SAMPLES
Description
AdaptiveResonance.jl convention for which 2-D dimension contains the number of samples.