按照模型的特殊性，定义了映射、邻域和连通的概念；
The concepts of mapping, neighborhood and connection are defined according to the characteristics of the model.
根据纹理特征的局部马尔可夫性和高斯变量的条件回归之间的关系，将复杂的模型选择转变为较简单的变量选择，应用惩罚正则化技巧同步选择邻域和估计参数。
The structure of the GGM is explored by the connection between the local Markov property of texture features and the conditional regression of Gaussian random variables.
不用多说，这将作为一系列新功能在Linux桌面和其他邻域的出现铺平道路。
Needless to say, this will pave the way toward a host of new capabilities on the Linux desktop and beyond.
应用推荐