隐马尔可夫模型

[zhongjiang1111]

您好，是这样的，我看到了您的工具箱，冒昧的打扰您一下，因为我最近在研究HMM 我遇到了下面问题 1.连续型隐马尔可夫（CHMM）中 我的观测序列是4维的，这个到底是多观测序列，还是一个观测序列啊 2.对于多维数据，进行kmeans聚类，以及求取前向后向算法，这个具体该怎么对应，能否向您请教请教 麻烦您了！ Hello, yes, i saw your toolbox, take the liberty of bothering you because i recently studied HMM I have encountered the following questions

1. The Continuous Hidden Markov (CHMM) in my observation sequence is 4-dimensional, this in the end is a multi-observation sequence, or an observation sequence ah
2. For multidimensional data, kmeans clustering, and seeking forward backward algorithm, the specific how the corresponding, whether to ask you to ask

[zhongjiang1111]

Hello, I am looking at your code, do not quite understand why ChmmGaussTest () when the data is generated Generate Data For i1 = 1: 2      X1 = mvnrnd ([0,0], [0.5, 0.2; 0.2, 0.3] / 5, 20);      X2 = mvnrnd ([0,2], [0.3, -0.2; -0.2, 0.5] / 5, 30);      X3 = mvnrnd ([0,4], [0.5, 0; 0, 0.3] / 5, 40);      X = [X1; X2; X3];      Data {i1} = X; End So ah

[Alwen-V]

您好，是这样的，我看到了您的工具箱，冒昧的打扰您一下，因为我最近在研究HMM 我遇到了下面问题 1.连续型隐马尔可夫（CHMM）中 我的观测序列是4维的，这个到底是多观测序列，还是一个观测序列啊 2.对于多维数据，进行kmeans聚类，以及求取前向后向算法，这个具体该怎么对应，能否向您请教请教 麻烦您了！ Hello, yes, i saw your toolbox, take the liberty of bothering you because i recently studied HMM I have encountered the following questions

1. The Continuous Hidden Markov (CHMM) in my observation sequence is 4-dimensional, this in the end is a multi-observation sequence, or an observation sequence ah
2. For multidimensional data, kmeans clustering, and seeking forward backward algorithm, the specific how the corresponding, whether to ask you to ask

你好！！我也有同样的问题！！请问您现在解决了嘛[help][help]

[qiuqiangkong]

您好，是这样的，我看到了您的工具箱，冒昧的打扰您一下，因为我最近在研究HMM 我遇到了下面问题 1.连续型隐马尔可夫（CHMM）中 我的观测序列是4维的，这个到底是多观测序列，还是一个观测序列啊 2.对于多维数据，进行kmeans聚类，以及求取前向后向算法，这个具体该怎么对应，能否向您请教请教 麻烦您了！ Hello, yes, i saw your toolbox, take the liberty of bothering you because i recently studied HMM I have encountered the following questions

1. The Continuous Hidden Markov (CHMM) in my observation sequence is 4-dimensional, this in the end is a multi-observation sequence, or an observation sequence ah
2. For multidimensional data, kmeans clustering, and seeking forward backward algorithm, the specific how the corresponding, whether to ask you to ask

你好！！我也有同样的问题！！请问您现在解决了嘛[help][help]

for i1 = 1:2 代表一共有两条数据。每一条数据是X，X是2 * 90的矩阵，其中2是数据维度，90是数据长度。 如果输入数据是4维，那么输入格式是4 * T就可以了。