deletions | additions       diff --git a/GMM.tex b/GMM.tex  index c0aa60a..9619662 100644  --- a/GMM.tex  +++ b/GMM.tex 
...
  GMM (Gaussian mixture modeling) method maximizes the likelihood of the data set using EM (expectation-maximization) method.  1. Assume thatour  data has unimodal distribution: $\textbf{x} \sim N(\mu, \sigma^2)$. Calculate $\mu$ and $\sigma^2$ 2. Assume that data has bimodal distribution:  $\textbf{x} \sim N(\mu_1, \mu_2, \sigma^2_1, \sigma^2_2, p)$(bimodal)  Initial guess: $\mu_1 = \mu - \sigma, \mu_2 = \mu + \sigma, \sigma^2_1 = \sigma^2_2 = \sigma^2, p = 0.5$