高斯混合模型GMM的C++实现

在高斯混合模型中需要使用概率更新参数的地方，程序中都简化成为了1处理，否则计算一个正态分布的概率还是挺花时间的。但是除了将概率换成1，其他地方还是严格按照公式的，大家可以仔细推导一下，就会看出其中的差异-



Gaussian mixture model GMM C++ implementation

　 In the Gaussian mixture model parameters need to use probability to update the place, the program has become a simplified treatment, or calculate the probability of a normal distribution was quite time-consuming. But in addition to the probability replaced by 1 or elsewhere in strict accordance with the formula, we can derive a closer look, you will see the differences