Exemple #1
0
 def log_z(self, parameter_type='regular', stop_gradient=False):
     if parameter_type == 'regular':
         pi = self.get_parameters('regular', stop_gradient=stop_gradient)
         eta = Stats.LogX(pi)
     else:
         eta = self.get_parameters('natural', stop_gradient=stop_gradient)[Stats.X]
     return T.logsumexp(eta, -1)
Exemple #2
0
    X,
    T.ones([batch_size]),
    T.ones([batch_size]),
])
x_stats = Gaussian.pack([
    T.outer(X, X),
    X,
])
theta_cmessage = q_theta.expected_sufficient_statistics()

num_batches = N / T.to_float(batch_size)
nat_scale = 10.0

parent_z = q_pi.expected_sufficient_statistics()[None]
new_z = T.einsum('iab,jab->ij', x_tmessage, theta_cmessage) + parent_z
q_z = Categorical(new_z - T.logsumexp(new_z, -1)[..., None],
                  parameter_type='natural')
p_z = Categorical(parent_z - T.logsumexp(parent_z, -1),
                  parameter_type='natural')
l_z = T.sum(kl_divergence(q_z, p_z))
z_pmessage = q_z.expected_sufficient_statistics()

pi_stats = T.sum(z_pmessage, 0)
parent_pi = p_pi.get_parameters('natural')
current_pi = q_pi.get_parameters('natural')
pi_gradient = nat_scale / N * (parent_pi + num_batches * pi_stats - current_pi)
l_pi = T.sum(kl_divergence(q_pi, p_pi))

theta_stats = T.einsum('ia,ibc->abc', z_pmessage, x_tmessage)
parent_theta = p_theta.get_parameters('natural')[None]
current_theta = q_theta.get_parameters('natural')
Exemple #3
0
new_pi = p_pi.get_parameters('natural') + T.sum(z_pmessage, 0)
parent_pi = p_pi.get_parameters('natural')
pi_update = T.assign(q_pi.get_parameters('natural'), new_pi)
l_pi = T.sum(kl_divergence(q_pi, p_pi))

new_theta = T.einsum('ia,ibc->abc', z_pmessage,
                     x_tmessage) + p_theta.get_parameters('natural')[None]
parent_theta = p_theta.get_parameters('natural')
theta_update = T.assign(q_theta.get_parameters('natural'), new_theta)
l_theta = T.sum(kl_divergence(q_theta, p_theta))

parent_z = q_pi.expected_sufficient_statistics()[None]
new_z = T.einsum('iab,jab->ij', x_tmessage,
                 theta_cmessage) + q_pi.expected_sufficient_statistics()[None]
new_z = new_z - T.logsumexp(new_z, -1)[..., None]
z_update = T.assign(q_z.get_parameters('natural'), new_z)
l_z = T.sum(kl_divergence(q_z, Categorical(parent_z,
                                           parameter_type='natural')))

x_param = T.einsum('ia,abc->ibc', q_z.expected_sufficient_statistics(),
                   q_theta.expected_sufficient_statistics())
q_x = Gaussian(x_param, parameter_type='natural')
l_x = T.sum(q_x.log_likelihood(X))

elbo = l_theta + l_pi + l_z + l_x
elbos = []

sess = T.interactive_session()

draw()