Skip to main content

Table 2 EDAs taxonomy

From: A review of estimation of distribution algorithms in bioinformatics

Statistical order Advantages Disadvantages Examples
Univariate Simplest and fastest Ignore feature dependencies PBIL (Baluja, 1994)
  Suited for high cardinality problems Bad performance for deceptive problems UMDA (Mühlenbein and Paaß, 1996)
  Scalable   cGA (Harik et al., 1999)
Bivariate (statistics of order two) Able to represent low order dependencies Possibly ignore some feature dependencies MIMIC (De Bonet et al., 1996)
  Suited for many problems Slower than univariate EDAs Dependency trees EDA (Baluja and Davies, 1997) BMDA (Pelikan and Mühlenbein, 1999)
  Graphically inquire the induced models   Tree-EDA/Mixture of distributions EDA (Santana et al., 1999)
Multivariate (statistics of order greater than two) Parameter learning (only interaction model parameters)
  Suited for problems with known underlying model Possibly ignore complex feature dependencies FDA (Mühlenbein et al., 1999)
   Higher memory requirements than bivariate Markov network-based EDA (Shakya and McCall, 2007)
  Structure+parameter learning (interaction model & parameters of the model)
  Maximum power of generalization Highest computation time EcGA (Harik et al., 1999)
  Flexibility to introduce user dependencies Highest memory requirements EBNA (Etxeberria and Larrañaga, 1999)
  Online study of the induced dependencies   BOA/hBOA (Pelikan et al., 1999, 2005)
    Dependency networks EDA (Gámez et al., 2007)
  1. A taxonomy of some representative EDAs. We highlight a set of characteristics that can guide the choice of a particular EDA suited to the goals and properties of a given problem.