A type of
neural network using kernel-based approximation to form an estimate of the probability density functions of classes in a
classification problem. One of the so-called Bayesian networks (Speckt, 1990; Patterson, 1996; Bishop, 1995). See also,
Neural Networks .