EM Mean Update 1D

Problem

Perform the 1D EM mean update for each component: new_mean_k = sum_i r_ik * x_i / sum_i r_ik, keeping old_means[k] when the responsibility sum is 0.

responsibilities is an n x K matrix. Empty old_means returns [].

Hint

The new mean is the responsibility-weighted average; an all-zero column leaves the old mean untouched.

case 1

in: [[1,2,3],[[1,0],[0.5,0.5],[0,1]],[0,0]]

expected: [1.3333333333333333,2.6666666666666665]

case 2

in: [[1,2],[[0,1],[0,1]],[5,5]]

expected: [5,1.5]

case 3

in: [[1],[[1]],[3]]

expected: [1]

case 4

in: [[],[],[]]

expected: []

Starter codePython

def em_mean_update(x, responsibilities, old_means):
    # Your code here
    pass

4 test cases · run your code against them in the editor