Extraction of developing type for construction of stochastic causality models from large scale historical data set

点击翻译按钮获取中文摘要

We propose some methodology to extract individual clusters and developing type's models for construction of stochastic causality models from large scale historical data set. For clustering individual set, we make some cases using classical self-organizing maps, and propose flexible self-organizing maps, as a new method to present intrinsic topological structure of individuals. And we make some prediction models to estimate on the clusters.