李群深层结构学习算法研究

针对数据的复杂性和语义深层关系,提出一种李群深层结构学习算法。主要包括:基于流形的深层结构分析方法、基于参数的李群半监督学习算法和基于线性的李群半监督学习算法,以及这些算法相融合的李群深层结构学习算法。该算法对连续语义间的深层关系有着重要的作用。实验结果显示,深度越深,该算法的效果越好。

Research on Lie Group Deep Structure Learning Algorithm

For the complexity of data and the deep relationship of semantic, a Lie Group deep structure learning algorithm is proposed. These mainly include deep structure analysis method based on the manifold, semi-supervised learning algorithm based on GL(n) and parameter Lie Group, as well as the integration of these algorithms relative to a Lie Group deep structure learning algorithm. The algorithm plays an important role in the deep relationship of continuous semantic. The experimental results show that the deeper structure is, the better effect of the algorithm is.