Methods, computer-readable media, and systems are provided for the detection of propagating phase gradients using model field theory of non-Gaussian mixtures. One embodiment of the invention is directed to a method for identifying phase cones in a data set. The method includes generating an initial set of values of parameters (ξA,ηA,tA) for a preset number of cones, initializing a covariance matrix with high values, evaluating the ratio of volume of a phase cone over total observed space-time volume, comparing the data set with a model, estimating a class probability density function, and generating a revised set of values of parameters (ξA,ηA,tA).
展开▼
机译:提供了用于使用非高斯混合物的模型场理论来检测传播相梯度的方法,计算机可读介质和系统。本发明的一个实施例涉及一种用于识别数据集中的相锥的方法。该方法包括生成用于预设数量的圆锥的参数(ξ A Sub>,η A Sub>,t A Sub>)的初始值集,进行初始化一个具有高值的协方差矩阵,评估相锥体积与总观测时空体积之比,将数据集与模型进行比较,估算类别概率密度函数,并生成一组修改后的参数值(ξ A Sub>,η A Sub>,t A Sub>)。
展开▼