Milestones

List view

• Include more probabilistic models to learn distributions

  Currently, only the variational Gaussian mixture model class is used in GrainLearning. We could replace GMM with any mixture of probability density functions. The idea is to have a "sampling method" class separated from the "distribution model" class (currently contains Halton sequence and Bayesian Gaussian mixture).

  No due date
  0% complete0 open0 closed