Aan de slagGa gratis aan de slag

Maximization function

We saw that the EM algorithm is an iterative method between two steps: the expectation and the maximization. In the last exercise, you created the expectation function. Now, create the maximization function which takes the data frame with the probabilities and outputs the estimations of the means and proportions.

Deze oefening maakt deel uit van de cursus

Mixture Models in R

Cursus bekijken

Oefeninstructies

Create the function maximization by completing the sample code.

Praktische interactieve oefening

Probeer deze oefening eens door deze voorbeeldcode in te vullen.

maximization <- function(___){
  means_estimates <- data_with_probs %>%
    summarise(mean_1 = sum(x * ___) / ___(prob_cluster1),
              mean_2 = sum(x * ___) / ___(prob_cluster2)) %>% 
    as.numeric()
  props_estimates <- data_with_probs %>% 
    summarise(proportion_1 = ___(prob_cluster1),
              proportion_2 = 1 - ___) %>% 
    as.numeric()
  list(means_estimates, props_estimates)   
}
Code bewerken en uitvoeren