ManyExpertDecisionTrees.jl

An implementation of Many Expert Decision Trees.

ManyExpertDecisionTrees.manify — Function

manify(dt::DecisionTree.Root, X::DataFrame, experts::UnionAll...)

Convert a DecisionTree.jl decision tree into a ManyExpertDecisionTree by attaching N membership functions per node, parameterized from subdivisions of X.

source

ManyExpertDecisionTrees.addexperts! — Function

addexperts!(medt::ManyExpertDecisionTree, X::DataFrame, experts::UnionAll...)

Add an arbitrary number of experts to the MEDT. Each expert's corresponding membership function will be parameterized from a different subdivision of X.

source

ManyExpertDecisionTrees.apply — Function

apply(tree::ManyExpertDecisionTree{T}, MXA::ManyExpertAlgebra, instance::AbstractVector{Float64}) where {T}

Given an instance, evaluate its membership degree to each class using the tnorms defined by the ManyExpertAlgebra.

source