I used GMM to cluster my data. It has 4 dimensions with 3 clusters. The GMM ran very well, and I got the mean and covariance matrix for each cluster.
The problem is I don't know which label belongs to which cluster. In addition, the cluster labels changed whenever I ran the iteration.
For example, in the first iteration, the cluster labels were 1, 2 and 3. In the second iteration, the cluster labels were 2,1 and 3, etc.
Can I fix the cluster labels? If yes, how do I do it?
Do you know the labels for each data sample? If yes, then you can build a GMM for each class.
Do you know the labels for each data sample? If yes, then you can build a GMM for each class.
Yes, I know the labels. To build a GMM for each class means I have to separate the data according to its label and make a GMM for each separated data? The means and covariance are simple, but what about the prior (the weight)?
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering