Quantitative Decision & Risk Analysis


IDI analysts (IDIers) have a foundation in quantitative analytic techniques that span the spectrum of descriptive statistical analysis to predictive modeling to prescriptive optimization of portfolios of systems.  IDIers have advanced degrees in a variety of scientific disciplines and social sciences, and are proficient in the application of many analytic methods using many different techniques and software tools.  This expertise enables IDIers to succeed in any client environment with standard office software or client-preferred specialized tools that model interdependencies among related systems, probabilistic relationships among events, time series sequencing of physical systems or decision processes, and support value measurement among seemingly disparate decision alternatives and outcomes.

Noteworthy applications: 

  • Numerous trade studies for the engineering design of national security systems
  • Cost/benefit and business case analyses for multi-million dollar defense system acquisition
  • Process modeling and process improvement for large government agency handling hundreds of millions of cases annually
  • Predictive analysis of deterrent effects in Blue versus Red conflict scenario
  • Cyber security risk analysis and countermeasure portfolio optimization
  • Creation of a framework for modeling adaptive adversaries for use in terrorism risk assessments
  • Sample of techniques: multi-objective decision analysis; influence diagrams; discrete event simulation; Monte Carlo simulation; integer, linear and non-linear programming; statistics – classical and Bayesian; spreadsheet modeling

 

Source, Site, Solution Selection

Applying MODA techniques and facilitated analysis of stakeholder preference, IDI’s analysts work closely with decision makers and stakeholders to clarify their objectives, desired outcomes, and preferences.  IDI’s approach scores alternatives and illuminates findings such as which alternative is best and by how much, in comparison to the others.  IDI’s approach also illuminates objectives that are most/least likely to influence the selection of the best solution and what enhancements most positively influence desired outcomes.  Frequently, this analysis motivates new ideas and innovative approaches to achieve better outcomes than initially perceived possible.

 

Portfolio Management

Portfolio management requires understanding what projects comprise the portfolio, what benefits they provide to the organization as a whole, what their schedules are and what they cost.  Included in portfolio management is a management workflow that keeps the data for these projects updated.  Resource allocation decisions then need to be supported so that projects can be refined and/or expanded in order for the organization to continually adjust the portfolio to the demands of a rapidly changing world. These decisions usually do not involve an either/or choice among different functions or programs-- they require an appreciation for the blended complementary contributions of the portfolio’s components, which are frequently non-linear.  IDIers help clients decide how much of their limited resources to devote to each program or area in order to get the most “bang for the buck” out of a limited budget.

Risk Assessment

Influence diagrams, fault trees, failure mode and effect analysis, and decision trees are good ways to structure and analyze many decision problems.  IDIers help clients clarify threats, vulnerabilities, and consequences regarding risks and then define and clarify countermeasures options that may mitigate these risks.  IDI has developed a number of assessment methods for quantifying the likelihood and impact of these uncertain events and conditions.  Once this quantitative information is available, IDI has implemented a number of approaches for calculating expected costs and benefits, and priorities.  Fault trees are used to understand the risk of a strategy or system failing, and to find the best way to reduce the risk.  Bayesian networks are very useful to defining probabilistic dependencies.  Below is a pictorial of our cyber risk tool that implements many of these ideas. IDI has conducted independent, third party reviews of terrorism risk assessments for biological, radiological and nuclear threats.

 

Predictive Analytics

Trying to make predictions/forecasts and then draw conclusions from a large amount and variety of information can be a difficult task.  IDI decision analysts help clients build models that use empirical data and subject matter expertise to describe probabilistic relationships among events, decisions, and external factors relevant to outcomes of interest.  The models offer an ability to dynamically update information and observe the changes in the likelihood of the outcomes, events, and factors.  IDI analysts are skilled at using ensemble-modeling techniques such as Bayesian aggregation, random forests, and random Bayesian forests to reach results that others have not been able to reach using traditional descriptive statistical techniques