Gaussian Mixture Model (GMM)

• is a type of Process
• is a type of mixture model that uses a combination of Gaussian (Normal) probability distributions. Each Gaussian distribution has a separately parametrized mean 𝜇^(𝑖) and covariance 𝛴^(𝑖)
• not to be confused with Process (MVN)

GMM - Univariate Bimodal Model (Example)

GMM - Regression / Learning Parameters

• see Gaussian Process Regression (GPR) - Kriging