My main objective is to predict the posterior probability of an individual belonging to one of the classes, using Bayes theorem. The information I have is:

  1. value of the data point
  2. mean and stdev of the 2 potential populations, that this data point could belong to (=the potential classes
  3. normal distributions drawn on those 2 classes

I'm using numpy to create a probability distribution over a given mean and stdev. Let's say that this distribution represents the average weight of a population. I have drawn distributions of the weights of 2 populations, such that the means differ considerably.

  • G1 = scipy.stats.norm(50, 1)
  • G2 = scipy.stats.norm(100,2)

I have a data point, suppose A = weight of one single individual. If I get probability of 'a' on one of the curves, using:

  • weightofA = 45 kg
  • probA_G1 = G1.pdf(weightofA)

what would this probability be? The likelihood or the posterior? If it's the likelihood, how will I get the prior? I only have 1 data point. How will I calculate the marginal, of the gaussians?

Thanks so much!



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.