Reference no: EM133296815

You have a large data set without labels and have decided to use a GMM to cluster the data. You set the number of components set to three and fit the model.After a number of iterations , the algorithm converges.

A. Does the convergence mean that there are three clusters in the data? Or , in other words,does the convergence mean that the optimal number of components is three? Motivate

B. You input a new datum(data point) to the model , and instead of returning a component label(i.e an integer), it returns a three element vector of decimal numbers.What does this three-element vector represent, and how do you convert it to a component label?

C. You decided to use cross validation to fit a better model.Describe in three steps how you would do.

D. What is the primary model parameter (hyperparameter) you optimise with cross validation?