Monte Carlo simulation random number generators

Monte Carlo simulation is a numerical experimentation technique to obtain the statistics of the output variables of a function, given the statistics of the input variables. In each experiment or trial, the values of the input random variables are sampled based on their distributions, and the output variables are calculated using the computational model. The generation of a set of random numbers is central to the technique, which can then be used to generate a random variable from a given distribution. The simulation can only be performed using computers due to the large number of trials required. 

Monte Carlo simulation uses computer programs called random number generators. A random number may be defined as a nmnber selected from tlie interval (0, 1) in such a way tliat tlie probabilities that the number comes from any two subintervals of equal lengtli are equal. For example, the probability tliat tlie number is in tlie subinter al (0.1, 0.3) is the same as the probability tliat tlie nmnber is in tlie subinterval (0.5, 0.7). Random numbers thus defined are observations on a random variable X having a uniform distribution on tlie interval (0, 1). Tliis means tliat tlie pdf of X is specified by... 

Suppose that using Monte Carlo simulation witli 10 simulated values of Ta and 10 simulated values of Tb, it is desired to estimate an average value of Ts. First, 20 random numbers are generated. Tliese are shown in columns 1 and 4 of Table 20.6.2. Regard each of the random numbers generated as the value of tlie cdf of a standard nonnal variable Z. Let Zi be tlie simulated value of Z corresponding to 0.10, tlie first random number in colunm 1. Then, since 0.10 is tlie value of tlie cdf for Z = Zi,... 

The obvious lesson to be taken away from this amusing example is that how well a net learns the desired associations depends almost entirely on how well the database of facts is defined. Just as Monte Carlo simulations in statistical mechanics may fall short of intended results if they are forced to rely upon poorly coded random number generators, so do backpropagating nets typically fail to ac hieve expected re.sults if the facts they are trained on are statistically corrupt. 

Monte Carlo simulation can involve several methods for using a pseudo-random number generator to simulate random values from the probability distribution of each model input. The conceptually simplest method is the inverse cumulative distribution function (CDF) method, in which each pseudo-random number represents a percentile of the CDF of the model input. The corresponding numerical value of the model input, or fractile, is then sampled and entered into the model for one iteration of the model. For a given model iteration, one random number is sampled in a similar way for all probabilistic inputs to the model. For example, if there are 10 inputs with probability distributions, there will be one random sample drawn from each of the 10 and entered into the model, to produce one estimate of the model output of interest. This process is repeated perhaps hundreds or thousands of times to arrive at many estimates of the model output. These estimates are used to describe an empirical CDF of the model output. From the empirical CDF, any statistic of interest can be inferred, such as a particular fractile, the mean, the variance and so on. However, in practice, the inverse CDF method is just one of several methods used by Monte Carlo simulation software in order to generate samples from model inputs. Others include the composition and the function of random variable methods (e.g. Ang Tang, 1984). However, the details of the random number generation process are typically contained within the chosen Monte Carlo simulation software and thus are not usually chosen by the user. 

The backbone of Monte Carlo simulation is the ability to generate random numbers because random numbers form the basis of the random draws from a probability distribution. Computers, because they are based on rules, algorithms, and mathematical operations, cannot generate truly random numbers. Instead, random numbers start from some point in the algorithm, called the seed, and proceed in a linear, predictable manner but when examined in the short term appear to be random. It... 

---