Quantum Monte Carlo diagonalization method as a variational calculation

Malvin H. Kalos, G.A. Mikhailov

Takashi Yanagisawa, Soh Koike, Yamaji K.

We propose a Monte Carlo method, which is a hybrid method of the quantum Monte Carlo method and variational Monte Carlo theory, to study the Hubbard model. The theory is based on the off-diagonal and the Gutzwiller type correlation factors which are taken into account by a Monte Carlo algorithm. In the 4x4 system our method is able to reproduce the exact results obtained by the diagonalization. An application is given to investigate the half-filled band case of two-dimensional square lattice. The energy is favorably compared with quantum Monte Carlo data.

HX Huang, ZX Cao, 曹泽星

Takashi Yanagisawa, Takeshi Nakanishi, Soh Koike +1 more

We investigate the two-dimensional Hubbard model by the variational Monte Carlo method. We consider off-diagonal wave functions where we improve the Gutzwiller wave function by multiplying off-diagonal factors. In the 4×4 system, we compare our results with the exact diagonalization. The variational energies are fairly close to the exact values. We show the results for the 10×10 Hubbard model.

Ryo Maezono, Hirofumi Watanabe, Shigenori Tanaka +2 more

The total energy of a small polypeptide system is calculated by combining the quantum Monte Carlo (QMC) and fragment molecular orbital (FMO) methods. Electronic correlation is taken into account using Slater–Jastrow wave functions and the variational quantum Monte Carlo (VMC) method. We calculate the energy of the whole system directly and by using the FMO method, finding that the combined QMC–FMO approach works very well.

Huang Hong