Duration: 4 full days
Minimum of 6 people and maximum of 15 people
This 4-day course will cover the core principles of quantitative risk analysis and the most important risk modeling principles, methods and techniques. The course will be taught using the R statistical language but the lessons apply equally well to other modeling environments. The focus of the course is on how to conduct accurate and effective quantitative risk analyses, including best practices of risk modeling, selecting the appropriate distribution, using data and expert opinion, and avoiding common mistakes. The course will also cover essential probability and statistics theory and various stochastic processes to provide the participants with a solid understanding of quantitative risk analysis
Who should attend
Anyone in business, government and science with an interest in quantitative risk analysis such as professionals needing to perform quantitative risk analysis in epidemiology, finance or operations, engineering, project risk analysis and researchers who are involved in risk analyses. Also, people who have experience in risk analysis using spreadsheets but wants to learn how to use a more sophisticated modeling environment such as R.
Participants are required to bring laptops loaded with R, and a pdf reader. R is an open-source freeware and can be downloaded free of charge from the R Project website. As R is updated constantly, please download the latest version before attending the course. Also, it is recommended participants use a code editor such as Tinn-R (Windows) to facilitate the display and storage of their code.
Prior experience using R or other simulation tools is not required.
All of EpiX Analytic's courses aim to help the participants understand risk analysis from the bottom-up, which is achieved through a relaxed, informal and interactive environment using plenty of examples and hands-on exercises where students apply and adapt what they have learned.
ModelAssist is a comprehensive risk analysis training reference and is free of cost. This reference tool provides an in-depth explanation of all of the risk analysis concepts, techniques and methods introduced in this course and greatly complements the course material. ModelAssist can be downloaded directly from our website here.
Introduction to risk analysis
- Background of risk analysis and risk management
- Risk analysis as a team effort
- Going from data to knowledge to a useful decision tool
- Dealing with the limits of current knowledge
Introduction to statistical descriptors in the context of risk analysis
- Mean, mode, standard deviation, skewness, kurtosis, percentiles
Introduction to probability theory
- The use of distributions: uncertainty, variability and inter-individual variability
- Probability concepts
- Graphical representations of risk events: Venn diagrams, fault trees and event trees
- A look at some simple probability distributions
Risk modeling in R
- Data structures used in simulation modeling
- Basic data manipulation and exploration
- Probability distributions in R and their differences with other software
Risk modeling in R (continued)
- Using loops and vectorized calculations for simulation
- Storing and retrieving simulation results
- Graphical exploration of simulation data
- Basic simulation analyses and diagnostics
Basics of risk modeling
- Monte Carlo simulation
- Calculation vs. simulation - the pros and cons of Monte Carlo
- Typical risk analysis results, their presentation and interpretation
- Practical problems to solve
- The most common probability distributions
Stochastic processes - the basis of risk analysis
- Binomial Process
- Binomial, beta, negative binomial and geometric distributions
- Imperfect tests, machine failures, risk events, etc.;
- Poisson Process
- Poisson, gamma, and exponential distributions
- Modelling insurance claims, accidents, random outbreaks, etc.
- Hypergeometric process
- Hypergeometric and inverse Hypergeometric distributions
- Survey results, prevalence estimate with imperfect diagnostic test, gambling etc.
- Practical problems to solve
Good practices in risk modelling
Common mistakes and how to prevent them
Introduction to analyzing and using data for risk analysis
- Statistical techniques
- Why we need uncertainty distributions not confidence intervals in risk analysis
- Creating uncertainty distributions with standard Classical Statistical tests
- t-tests, z-tests, Chi-squared tests
- Examples of estimation of population mean and standard deviation
- The Bootstrap to include uncertainty
- The use of Bayesian Statistics in risk analysis
Example risk analyses (a range of examples will also be covered during the course).
Wrap up and review of course material.