May 25, 2024
 HELP 2023-2024 Graduate Academic Catalog [ARCHIVED CATALOG] Print-Friendly Page (opens a new window) Add to Portfolio (opens a new window)

# CVE 5811 - Applied Statistics and Modeling

3 lecture hours 0 lab hours 3 credits
Course Description
This course covers topics in statistics needed for the statistical analyses of architectural and civil engineering systems.  It also presents methods for developing statistical models.  Specific topics include: (1) determining if significant difference or equivalence exists between data sets using parametric and non-parametric methods, (2) experimental design, (3) constructing linear and non-linear regression models, (4) developing Monte Carlo models, (5) analyzing time-series, and (6) Bayesian statistics. (prereq: MTH 2480) (quarter system prereq: MA 262)
Course Learning Outcomes
Upon successful completion of this course, the student will be able to:
• Compare two or more treatments or data sets for differences using the t-test, Tukey HSD or non-parametric tests
• Evaluate the effects of one or more variables on the response of a system using full and fractional factorial designs
• Design testing programs to control the probability of Type I and Type II errors
• Build mathematical models of systems using linear and non-linear regression analysis, and time series analysis and estimate confidence intervals for model parameters
• Design experiments using full and partial factorial models and analyze the results of such experiments
• Compose artificial data sets suitable for modeling purposes using Monte Carlo simulation
• Utilize a commercial statistical package with facility to perform statistical analysis

Prerequisites by Topic
• None

Course Topics
• Review of statistics and definitions, data visualization, estimating percentiles, software introduction and use
• Performing hypothesis testing using parametric tests: “proving” difference or equivalence; understanding Type I and Type II error
• Non-parametric testing for “proving” difference; experimental design: design testing programs to control the probability of Type I and Type II errors
• Building models using linear regression, parsimony; transformations, problems with linearization
• Experimental design: measuring the effects of variables on an outcome using full and fractional factorial designs
• Time series analysis, auto and partial auto-correlations, smoothing
• Identifying distributions, Monte Carlo simulation
• Signal versus noise, Bayes theorem

Coordinator
Dr. William Gonwa, P.E.

Add to Portfolio (opens a new window)