Hierarchical mixtures of experts and the EM algorithm
Neural Computation
Varieties of Helmholtz machine
Neural Networks - 1996 Special issue: four major hypotheses in neuroscience
Bayesian Ying-Yang machine, clustering and number of clusters
Pattern Recognition Letters - special issue on pattern recognition in practice V
Adaptive mixtures of local experts
Neural Computation
Competition and multiple cause models
Neural Computation
Temporal BYY learning for state space approach, hidden Markovmodel, and blind source separation
IEEE Transactions on Signal Processing
| Hi-index | 0.00 |
Bayesian Ying-Yang (BYY) learning is proposed as a unified statistical learning framework firstly in (Xu, 1995) and systematically developed in past years. Its consists of a general BYY system and a fundamental harmony learning principle as a unified guide for developing new parameter learning algorithms, new regularization techniques, new model selection criteria, as well as a new learning approach that implements parameter learning with model selection made automatically during learning (Xu, 1999a&b; 2000a&b). This paper goes further beyond the scope of BYY learning, and provides new results and new understandings on harmony learning from perspectives of conventional parametric models, BYY systems and some general properties of information geometry.