Data Science Stack Exchange
Data Science Stack Exchange

Questions tagged [gmm]

For questions about Gaussian mixture models (GMMs).

10 questions
3
votes
1 answer

Why Gaussian mixture model uses Expectation maximization instead of Gradient descent?

Why Gaussian mixture model uses Expectation maximization instead of Gradient descent? What other models uses Expectation maximization to find best optimal parameters instead of using gradient descent?
star
  • 1,411
  • 7
  • 18
  • 29
2
votes
1 answer

Which algorithm or tool to use to classify as good or bad?

I have a feature vector with different data types, Considering all the data in that feature vector. I have to classify as Good or Bad. Which algorithm should be used to just get the output Good or bad based on different data types in a feature…
Amruth
  • 29
  • 2
2
votes
1 answer

Math of Gaussian Mixture Models & EM Algorithm

I'm having a hard time understanding the math behind GMMs. A GMM is a weighted sum over K different Gaussian components with parameters $\mu_k, \sigma_k, \pi_k$ From my understanding, the general overview is: Initialize random parameters for each…
bosonphoton
  • 23
  • 3
1
vote
1 answer

Gaussian Mixture Models Clustering

When using the EM algorithm in Gaussian Mixture Models (GMM), in the E-step, we take each x set in the training dataset to calculate and update the "weight" and parameters of each Gaussian distribution of the clusters (M-step). I have read that we…
YCCCCC
  • 111
  • 5
1
vote
1 answer

Treating Word Embeddings as Multivariate Gaussian Random Variables

I want to specify some probabilistic clustering model (such as a mixture model or lda) over words, and instead of using the traditional method of representing words as an indicator vector , I want to use the corresponding word embeddings extracted…
ricardo
  • 23
  • 5
0
votes
0 answers

scikit GMM fails randomly with ill-defined linear covariances

I am trying to fit a GMM model to my data. The dataset has 37 features (some int and others float). When the dataset has small number of rows (<200), GMM fits OK. When I try a larger dataset (500 rows), some times (2/15), GMM fails to fit and gives…
RedBaron
  • 101
  • 2
0
votes
1 answer

Method to Compare the Fit of of k-Medoids and GMM to my Dataset

I'm looking for a method(s) to compare the fit of k-Medoids and a GMM. Currently, I'm looking at the distance between the max-min means of the GMM clusters and comparing that to the max-min medoid locations of the k-Medoids. Is this a good method or…
Andrew Feenan
  • 1
  • 1
0
votes
0 answers

Expectation Step in Gaussian Mixture Model for Matrix Data Not Producing Proper Posterior Probabilities

I'm working on implementing a Gaussian Mixture Model (GMM) for three-way data (i.e., a set of matrices) in R. The GMM is being estimated using the Expectation-Maximization (EM) algorithm. However, I'm encountering an issue during the Expectation (E)…
Omarov Alen
  • 1
0
votes
0 answers

Fit a bimodal histogram with a mixture Model that is not Gaussian

I am attempting to fit with Python a histogram which presents a bimodal distribution. For now, I have tested to use the Gaussian Mixture Model (GMM) from Scikit-Learn, but I want to try with different combinations of distributions that are not…
Jack
  • 1
0
votes
0 answers

Can you learn about Hidden Markov Models with Gaussian Mixture Models?

I am using both somewhat concurrently: I notice that the both have and entanglement of the number of features and number of components. Also, both describe covariance matrices in the same way.
Thunder
  • 11
  • 1