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.
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。
Kernal methods and support Vector Machines (SVMs) are related to Gaussian processes and can also be used in classification and regression problems.
内核方法和支持向量机(SVMs)与高斯过程相关并应用于分类和回归问题。
应用推荐