各位大大好,
我用python hmmlearn package
使用GaussianHmm时
model = hmm.GaussianHMM(n_components=10, covariance_type='full',min_covar
=0.00001, tol=0.0001, n_iter=NUM_ITERS, init_params='stmc')
model.fit(train_dataset)
程式执行有时会有错误讯息(有时正常执行) 如下:
component 3 of 'full' covars must be symmetric, positive-definite
修改covariance_type='spherical' 也会有错误讯息 如下:
'spherical' covars must be non-negative
请问有人知道这个问题如何解决吗
非常谢谢!