Skip to main content
Fig. 1 | Earth, Planets and Space

Fig. 1

From: Comparison between the Hamiltonian Monte Carlo method and the Metropolis–Hastings method for coseismic fault model estimation

Fig. 1

Example of sampling using the Hamiltonian Monte Carlo (HMC) method. This figure shows the exploration of a mean parameter of a simple normal distribution from synthetic gaussian data \(\left( {f\left( {\theta {\text{|d}}} \right) \propto {\text{exp}}\left( { - \theta^{2} } \right)} \right)\). The contours show the constant Hamiltonian and their values. Triangles, circles, and squares show the start points, transition points, and end points of each leapfrog transition, respectively. The blue triangle particularly represents the initial point of the sampling. In addition, the orange line shows the jump to the next momentum generated by standard normal distribution after acceptance of the candidate point

Back to article page