Decision Support Tools and Analytics Research

IDI is often asked by clients to develop decision support tools for their use during day-to-day operations.  In addition, IDI has been an active and important contributor to research teams on very challenging problems.  In some of these activities, IDIers are able to prototype a set of analytical methods that have been fully researched by a team of government and IDI analysts.  In other cases, IDIers assemble a set of tools based on their tasking and present the tools to the client for use.  Still in other cases, IDI is tasked to create an analytical set of methods and tools implementing those methods for delivery to the client.  Finally, IDI is also able to create innovative decision support tools based on analytics methods developed and refined at IDI.

Cyber Risk Tool

 This is an overlay always on

Innovative Decisions, Inc. (IDI) created a Cyber Risk Tool that provides estimates of the risk level to a system or network. These risk estimates are based upon (1) information provided by the system administrator, risk officer or designer about the system, (2) information about what security controls and countermeasures are (or may be) in place, (3) designated threats to the system, and (4) a model of attacks, which consists of aggregate attack steps that are assembled into complete attacks. Preference theory and graph theory were used in developing the adversary and attack models with a Multiple Objective Decision Analysis approach to assessing risk.

The output of the Cyber Risk Tool is a set of heat maps for Confidentiality, Integrity, and Availability. A drill-down capability is provided with each heat map to determine which attacks and attack steps have the highest risk (probability of success paired with impact given success), as well as which controls and countermeasures, if implemented, would reduce these risks the most.


POM Pricing Tool

 This is an overlay on hover

IDI provided Navy Installation Command with a tool and process to enhance public safety managers’ abilities to quantify physical resources (e.g., personnel, equipment, facilities, etc.) required to meet performance standards; estimate the financial resources needed to procure and sustain the physical resources; and to communicate these needs to a wide variety of audiences.  IDI has helped the client examine programmatic opportunities and risks at a much more detailed level, while also supporting assessment of program execution against program plans and identification of program priorities with limited resources.  IDI’s efforts have facilitated understanding of program needs and priorities; improved the accuracy and timeliness of program resource calculations; and illuminated critical considerations during dynamic operational and budgetary environments.

Integrated Crisis Early Warning System

On a team with Lockheed Martin - Advanced Technology Laboratory, IDI integrated computational social science models to forecast country instability for 29 countries. The forecasters used reports from open sources coded by event type along with political science variables to produce forecasts. IDI’s Bayesian EOI Aggregator combined the forecasts into a set of predictions for five types of instability events of interest (EOIs): Ethnic-Religious Violence, Domestic Political Crisis, Insurgency, Rebellion and International Crisis. The IDI Bayesian EOI Aggregator used Bayesian network structure learning with political science variables to produce meaningful priors and employed a naïve Bayes structure to aggregate the forecasts.


Intelligent Non-Player Characters in Virtual Training Environments


IDI is currently developing automated BOTs for use as non-player characters (NPCs) in virtual training environments. The BOTs are implemented using a technology called Dynamic Decision Networks (DDNs), which use Bayesian networks to integrate inputs to the NPC from the player and influence diagrams to guide NPC decisions in order to adapt the NPC decisions as the training scenario unfolds.  In the current implementation, players send text statements and questions to the DDN-based NPC. The NPC then determines how warmly and truthfully to respond and provides the appropriate textual response. These responses range from helpful statements of truth to outright lies or ending the conversation. For more information on DDNs please see the DDN Overview.