EM
Project for CS5339 python implementation of Expectation Maximization Algorithm
Update parameters
ground truth:mu_1:(3.00, 3.00), mu_2:(-3.00, -3.00), var_1:(2.00, 4.50), var_2:(1.00, 8.00), lambda:(0.4, 0.6)
iteration:1.00, update:43.97, mu_1:(-4.00, 5.00), mu_2:(4.00, -5.00), var_1:(1.00, 1.00), var_2:(1.00, 1.00), lambda:(0.70, 0.30)
iteration:2.00, update:12.79, mu_1:(-0.53, 1.56), mu_2:(-0.71, -2.78), var_1:(10.10, 7.84), var_2:(9.79, 10.96), lambda:(0.47, 0.53)
iteration:3.00, update:13.49, mu_1:(0.97, 1.69), mu_2:(-2.01, -2.87), var_1:(9.73, 7.71), var_2:(5.99, 10.18), lambda:(0.47, 0.53)
iteration:4.00, update:7.85, mu_1:(1.94, 2.25), mu_2:(-2.76, -3.24), var_1:(6.93, 5.73), var_2:(2.38, 7.54), lambda:(0.45, 0.55)
iteration:5.00, update:2.89, mu_1:(2.56, 2.63), mu_2:(-3.02, -3.28), var_1:(3.90, 4.18), var_2:(1.12, 6.73), lambda:(0.43, 0.57)
iteration:6.00, update:0.87, mu_1:(2.88, 2.74), mu_2:(-3.04, -3.14), var_1:(2.14, 3.95), var_2:(1.05, 7.14), lambda:(0.41, 0.59)
iteration:7.00, update:0.10, mu_1:(2.98, 2.77), mu_2:(-3.02, -3.09), var_1:(1.67, 3.90), var_2:(1.06, 7.33), lambda:(0.40, 0.60)
iteration:8.00, update:0.01, mu_1:(2.99, 2.78), mu_2:(-3.02, -3.08), var_1:(1.63, 3.89), var_2:(1.07, 7.35), lambda:(0.40, 0.60)
Result
reference:
matlab code https://github.com/rezaahmadzadeh/Expectation-Maximization