A process for anticipating and shaping adversarial behavior

Abstract

Abstract:A new approach for anticipating and shaping adversarial behavior is developed and demonstrated. The approach extends the notion of prediction, which is a forecast of the future from a third party point of view, to anticipation, which is a forecast from the perspective of an entity having partial control in a domain. Shaping utilizes the models developed for anticipation to determine actions that influence another influential entity (e.g., an enemy) and actions to direct the emergent phenomena of a domain according to an entity's objectives. The approach is developed using principles of control theory and demonstrated in the southeastern region of Afghanistan. A key capability demonstrated by this approach is its ability to handle proactive adversaries when actionable intelligence is nonexistent. In the demonstration, Taliban (Red) combat actions are anticipated from the perspective of the coalition forces (Blue) across time, across space, and by the current state of the region and then shaped to Blue's desires. Shaping identifies periods of time that simultaneous or alternating Blue combat actions in different regions help meet Blue's military and nonmilitary objectives.

Some thoughts on the role of robust control theory in climate-related decision support

“…There are literally hundreds of different control algorithms that may have potential use in climate policy studies. For example, McKay et al (2011) utilized model predictive control with a minimax criteria to identify robust strategies for coalition forces in Afghanistan under the deep uncertainty caused by the insurgency. This application represented uncertainty, not with a single parameter representing a set of alternative models, but rather with a set of plausible scenarios derived from historical data.…”

Getting Real About Virtual Worlds

“…Based on Wand and Weber's (1994) framework of information structure, the design principles are partitioned into two categories: deep structure (DS) for the modeling and simulation dimension, and emergent structure (ES) for the emergent knowledge process dimension. Sentient World Simulation (SWS), which is a "large-scale multi-agent model consisting of over 12 million heterogeneous, intelligent agents representing the virtual populations, leaders, organizations, institutions, infrastructure, and geographies of over 60 nations" (McKay, Chaturvedi, & Adams, 2011), is used to illustrate the use and implementation of the agent-based virtual world design principles.…”

Using Real-Time Decision Tools to Improve Distributed Decision-Making Capabilities in High-Magnitude Crisis Situations*

“…Prior related research on collaborative decision making in crisis response has examined a wide variety of issues including goal competition and conflict (Kumar & Van Dissel, 1996; Osborn & Hagedoorn, 1997; Westley & Vrendenburg, 1997; Hardy & Phillips, 1998; Kumar & Niti, 1998; Drnevich et al, 2009), decision‐making effectiveness (Lin & Carley, 1993, 1997; Drnevich et al, 2010), response strategies (Carley et al, 2004; Green & Kolesar, 2004), and behavior anticipation and shaping (McKay et al, 2011). One common thread in this body of work is that multi‐organizational collaborative decision making is a multifaceted issue, and research on this topic may take a variety of forms.…”

“…As such, anticipation of behaviors of decision makers at different levels becomes important for an effective response. For example, decision makers in the federal government need to anticipate the response at the state and the local levels of government; likewise the state and local level decision makers need to anticipate decisions at multiple levels to inform their decisions (McKay, Chaturvedi, & Adams, 2011).…”