INPUT MODELLING
NAME: SOHAIL SHAIKH
ROLL NO. : PA03
SUB: DESS
CAD/CAM/CAE
CONTENT
• Purpose & Overview
• Selecting the Family of
• Data Collection Distributions
• Identifying the Distribution • Goodness-of-Fit Tests
• Histograms • Summary
PURPOSE & OVERVIEW
• Input models provide the driving force for a simulation model.
• The quality of the output is no better than the quality of inputs.
• In this chapter, we will discuss the 4 steps of input model development:
a. Collect data from the real system
b. Identify a probability distribution to represent the input process
c. Choose parameters for the distribution.
d. Evaluate the chosen distribution and parameters for goodness of fit.
DATA COLLECTION
• One of the biggest tasks in solving a real problem. GIGO – garbage-in-garbage-out.
• Suggestions that may enhance and facilitate data collection:
a. Plan: begin by a practice or pre-observing session, watch for unusual circumstances
b. Analyze the data as it is being collected: check adequacy.
c. Combine homogeneous data sets, e.g. successive time periods, during the same time period on successive days.
d. Be aware of data censoring: the quantity is not observed in its entirety, danger of leaving out long process times.
e. Check for relationship between variables, e.g. build scatter diagram.
f. Check for autocorrelation.
g. Collect input data, not performance data.
IDENTIFYING THE DISTRIBUTION
• Histograms
• Selecting families of distribution
• Parameter estimation
• Goodness-of-fit tests
• Fitting a non-stationary process
HISTOGRAMS
• A frequency distribution or histogram is useful in determining the shape of a distribution.
• The number of class intervals depends on:
a. The number of observations.
b. The dispersion of the data.
c. Suggested: the square root of the sample size.
• For continuous data:
a. Corresponds to the probability density function of a theoretical distribution.
• For discrete data:
a. Corresponds to the probability mass function
• If few data points are available: combine
adjacent cells to eliminate the ragged
appearance of the histogram.
• Vehicle Arrival Example: of vehicles arriving at
an intersection between 7 am and 7:05 am was
monitored for 100 random workdays.
• There are sample data, so the histogram may
have a cell for each possible value in the data
range
SELECTING THE FAMILY OF DISTRIBUTIONS
• A family of distributions is selected based on:
a. The context of the input variable.
b. Shape of the histogram.
• Frequently encountered distributions:
a. Easier to analyze exponential, normal and Poisson.
b. Harder to analyze: beta, gamma and Weibull
• Use the physical basis of the distribution as a guide, for example:
a. Binomial: of successes in n trials.
b. Poisson: of independent events that occur in a fixed amount of time or space.
c. Normal: dist’n of a process that is the sum of a number of component processes.
d. Exponential: time between independent events, or a process time that is memoryless
e. Weibull: time to failure for components.
f. Discrete or continuous uniform: models complete uncertainty.
g. Triangular: a process for which only the minimum, most likely, and maximum values are known
h. Empirical: resamples from the actual data collected.
• Is the process naturally discrete or continuous valued.
• Goal: obtain a good approximation
QUANTILE-QUANTILE PLOTS
• Q-Q plot is a useful tool for evaluating distribution fit
• If X is a random variable with cdf F, then the q-quantile of X is the γ such that.
• When F has an inverse, γ = F-1(q)
• Let {xi, i = 1,2, …., n} be a sample of data from X and {yj, j = 1,2, …, n} be the
observations in ascending order:
GOODNESS-OF-FIT TESTS
• Conduct hypothesis testing on input data distribution using:
a. Kolmogorov-Smirnov test
b. Chi-square test
• No single correct distribution in a real application exists.
a. If very little data are available, it is unlikely to reject any candidate distributions.
b. If a lot of data are available, it is likely to reject all candidate distributions.
SUMMARY
• In this chapter, we described the 4 steps in developing input data models:
a. Collecting the raw data.
b. Identifying the underlying statistical distribution.
c. Estimating the parameters
d. Testing for goodness of fit
Thank You
