History Of Monte Carlo Method

The current physics research can be divided into three fields of theory, experiment, calculation. Numerical methods of simulating the system as accurately as possible using a computer, constructing a computer model and providing a well-behaved experimental system, for example, the bridge between theory and experiments is increasing more and more . Kinetic method (MD). In the Monte Carlo method, the exact dynamic behavior of the system is replaced by a stochastic process, and the MD method is based on a simpler principle including solving the Newton equation of the N body system.

The Markov chain Monte Carlo method has two parts. Monte Carlo is a common method of using repeated sample at random to obtain numerical answers. Monte Carlo can be thought of as doing a lot of experiments, changing the variables in the model, and observing the response each time. By choosing random values ​​you can search most of the parameter space which is the range of possible values ​​of the variable. The parameter space of the problem using normal a priori variables (details are described below) are as follows.

This name refers to the famous casino of Monaco. It was created in 1949 by Stanislaw Ulam, one of the pioneers of this method. According to reports, urum's uncle is a gambling person, and the relationship between the element of 'opportunity' of gambling and the Monte Carlo law is particularly evident for Stanislav. Endresult = @ spawn hello ("World!") Print (result) fetched = fetch (result) print (fetched) The @everywhere macro guarantees that the hello () function is defined in all processes. The @ spawn macro is used to wrap the closure around the expression hello ("World!") and to automatically evaluate it remotely in the chosen process.

Monte Carlo / Simulation JWT contributed to Monte Carlo's technology. In 1956, Trotter and Tukey introduced the conditional Monte Carlo method. In this method, we estimated the integral of the weight function by a simple Monte Carlo rather than the integral of the function itself. This is one way to reduce variance. Among them, the sample space is embedded in a larger sample space. In the lecture at Princeton University in 1963, when you want to perform discrete Fourier transform on sequences of N sequences, when N = GH, you need (H + 2 + G) GH It is only multiplication. . He further pointed out that for N = 4k, less than 2N + N log 2 N is needed. The resulting Cooley-Tukey paper has become a standard reference. As FFT is promoted, signal processing immediately switches from analog to digital