R2GenOpt is a univariate discretization algorithm that we introduced in version 6.0. This algorithm uses a genetic algorithm to find a discretization that maximizes the R^{2} between the discretized variable and its corresponding continuous (hidden) variable. Therefore, it is the the optimal approach for finding an accurate representation of the continuous values of a variable.
As of version 8.0, BayesiaLab offers an extended version of R2GenOpt that uses a specific MDL score to choose the number of bins.
