126
6 Unsupervised Deep Learning
In this way, the desired convergence is guaranteed even if the probability defined by
H eff
θ does not exactly match the target P . Note that, to implement this correction
term, one needs to have access to the value of P (x). For example, if the target is
written by some Hamiltonian
P (x) =
e −H true (x)
Z
,
(6.115)
it is possible. In this case, the training sample is a snapshot of the configuration from
the statistical mechanical system defined by this Hamiltonian. A machine learning
Metropolis method that repeats the cycle of (1) training H
eff
θ with the configurations
generated by Markov chain Monte Carlo method for (6.115), and (2) generating new
configurations with a Markov chain of type (6.114), is called a self-learning Monte
Carlo method, which has been actively studied since 2016 [101].
6 Unsupervised Deep Learning
In this way, the desired convergence is guaranteed even if the probability defined by
H eff
θ does not exactly match the target P . Note that, to implement this correction
term, one needs to have access to the value of P (x). For example, if the target is
written by some Hamiltonian
P (x) =
e −H true (x)
Z
,
(6.115)
it is possible. In this case, the training sample is a snapshot of the configuration from
the statistical mechanical system defined by this Hamiltonian. A machine learning
Metropolis method that repeats the cycle of (1) training H
eff
θ with the configurations
generated by Markov chain Monte Carlo method for (6.115), and (2) generating new
configurations with a Markov chain of type (6.114), is called a self-learning Monte
Carlo method, which has been actively studied since 2016 [101].