This article explores the utility of neural network ensembles in knowledge discovery and integration. A novel neural network ensemble model KBNNE (Knowledge-Based Neural Network E...
dely acknowledged that categorically organized abstract knowledge plays a significant role in high-order human cognition. Yet, there are many unknown issues about the nature of ho...
In this article we present an incremental method for building a mixture model. Given the desired number of clusters K ≥ 2, we start with a two-component mixture and we optimize t...
Inference is a key component in learning probabilistic models from partially observable data. When learning temporal models, each of the many inference phases requires a complete ...
This study shows that a mixture of RNN experts model can acquire the ability to generate sequences that are combination of multiple primitive patterns by means of self-organizing ...