I came across the following link: Bayesian recipes (3): Simple, efficient statistics for Normal distributions in the link, I saw the part where the author says:
Focusing on the mean:
This Bayesian recipe is a little more complicated than some others in this series because the frequency distribution has two parameters (the mean and standard deviation) rather than just one. The Bayesian calculations give a surface showing how likely you think each possible combination of mean and standard deviation is, given the data and your initial beliefs. However, this is often just too much information! Usually it is the mean that we care about most. To understand how our beliefs about the true mean should be shaped we need to somehow sum all the possibilities over all the potential standard deviations. Fortunately, there is a well known distribution that does just that. It’s called Student’s T. Student’s T distribution has three parameters, called df (degrees of freedom), mu, and sigma. These can be calculated from the parameters of the normal-gamma using these simple formulae:
df = 2 × alpha
mu = mu (yes that’s right, it’s the same number for both distributions
sigma = beta / (alpha × kappa)
I would like to know if it’s possible to use equations for mu and sigma? I have been looking for articles for getting mu, sigma, beta and alpha.