Print

Quantitative Risk Assessment Methods: Part 2 - SIP

Register

Sign up for this course now, by clicking here.

Course Description

This 2.5 day face-to-face course provides participants with the opportunity to develop, scrutinize and present Monte Carlo simulation models.

Building on knowledge from Part I, Quantitative Risk Assessment Methods: Part 2 provides students with a strong foundation in stochastic processes, probabilistic risk assessment and Monte Carlo simulation. Students will gain a deeper understanding of the principles and mechanics of Monte Carlo simulation, build models using these principles, and learn how to analyze probabilistic models in a risk assessment context. We will also discuss how to use data and expert opinion when building models. Participants can expect to gain hands-on experience in building and analyzing computer-based probabilistic models. They will also experience some techniques and challenges to expect in presenting their results to various audiences.

Learning by example, students will be given exercises involving elements of real world risk assessments that are used in current policy and risk management problems.

The course is conducted in a computer teaching laboratory with two instructors. Lectures will describe and demonstrate various techniques. Students then work individually and in groups to solidify their understanding of the lecture materials, and to build quantitative modeling and analysis skills.

We strongly recommend Food Safety Risk Assessment and Quantitative Risk Assessment Methods: Part 1 as prerequisites to this course.

Overview of Topics

More Insight into Commonly-Used Distributions

  • The Normal and Lognormal Distributions and the Central Limit Theorem
  • The Uniform and Triangular Distributions
  • The Beta, Gamma and Exponential Distributions
  • Common Discrete Distributions: Bernoulli, Binomial, Poisson
  • Inter-Relationships, Approximations, and Mixtures
  • Exercise: Exploring Alternative Distributions using @RISK™

Choosing and Justifying Distributions

  • Choosing distributions given measured data
  • Choosing distributions on theoretical grounds
  • Choosing distributions given expert opinion
  • Truncating distributions
  • Correlating distributions
  • Exercises:
    • Fitting a distribution to data
    • Eliciting a distribution from an expert
    • The impact of correlated variables

Understanding Monte Carlo Simulation

  • Understanding convergence in simulation statistics
  • Sampling methods (Random sampling, Latin Hypercube sampling)
  • Understanding simulation problems for systems with rare events
  • Alternatives to 'brute force' computation

Advanced Concepts in Simulation

  • Separating uncertainty and variability
  • What is a 2-dimensional simulation model?

Model Building Exercises

  • Converting a system description into a probabilistic model
  • Implementing a probabilistic model in Excel
  • Analyzing the model using Sensitivity and Importance Analysis
  • Transparency: Documenting a model

Exercise: Presenting and explaining a probabilistic model

Learning Objectives

After completing this course, students understand:

  • Simulation principles and techniques
  • Stochastic processes
  • Scenario and Sensitivity analysis
  • How to use data when building a model
  • How to present risk assessments and their results