A5073745651- Cognitive Science and Mapping
- AI-based Problem Solving and Planning
- Opinion Dynamics and Social Influence
- Human-Automation Interaction and Safety
- Neural dynamics and brain function
- Neural and Behavioral Psychology Studies
- Complex Systems and Decision Making
- Information and Cyber Security
- Complex Network Analysis Techniques
- EEG and Brain-Computer Interfaces
- Semantic Web and Ontologies
- Intelligent Tutoring Systems and Adaptive Learning
- Neural Networks and Applications
- Explainable Artificial Intelligence (XAI)
- Reinforcement Learning in Robotics
- Bayesian Modeling and Causal Inference
- Computability, Logic, AI Algorithms
- Cognitive Computing and Networks
- Game Theory and Applications
- Cognitive Science and Education Research
- Evolutionary Game Theory and Cooperation
- Decision-Making and Behavioral Economics
- Child and Animal Learning Development
- Functional Brain Connectivity Studies
- Language and cultural evolution
Carnegie Mellon University
2015-2024
Decision Sciences (United States)
2003-2016
Alion Science and Technology (United States)
2010
University of Waterloo
2009
Human Computer Interaction (Switzerland)
2003
Adaptive control of thought-rational (ACT-R; J. R. Anderson & C. Lebiere, 1998) has evolved into a theory that consists multiple modules but also explains how these are integrated to produce coherent cognition. The perceptual-motor modules, the goal module, and declarative memory module presented as examples specialized systems in ACT-R. These associated with distinct cortical regions. place chunks buffers where they can be detected by production system responds patterns information buffers....
Abstract This paper presents a learning theory pertinent to dynamic decision making (DDM) called instancebased (IBLT). IBLT proposes five mechanisms in the context of decision‐making process: instance‐based knowledge, recognition‐based retrieval, adaptive strategies, necessity‐based choice, and feedback updates. suggests DDM people learn with accumulation refinement instances, containing situation, action, utility decisions. As makers interact task, they recognize situation according its...
Abstract The ACT-R system is a general for modeling wide range of higher level cognitive processes. Recently, it has been embellished with theory how its processes interact visual interface. This includes attention can move across the screen, encoding information into form that be processed by ACT-R. applied to several classic phenomena in literature depend on speed and selectivity which display. capable interacting same computer screens subjects do and, as such, well suited provide model...
The basal ganglia play a central role in cognition and are involved such general functions as action selection reinforcement learning. Here, we present model exploring the hypothesis that implement conditional information-routing system. system directs transmission of cortical signals between pairs regions by manipulating separately sources destinations information transfers. We suggest mechanism provides an account for several cognitive ganglia. also incorporates possible which subsequent...
Abstract Erev, Ert, and Roth organized three choice prediction competitions focused on related tasks: One shot decisions from description (decisions under risk), one experience, repeated experience. Each competition was based two experimental datasets: An estimation dataset, a dataset. The studies that generated the datasets used same methods subject pool, examined decision problems randomly selected distribution. After collecting data to be for estimation, organizers posted them Web,...
A standard model captures a community consensus over coherent region of science, serving as cumulative reference point for the field that can provide guidance both research and applications, while also focusing efforts to extend or revise it. Here we propose developing such humanlike minds, computational entities whose structures processes are substantially similar those found in human cognition. Our hypothesis is cognitive architectures appropriate abstraction defining model, although not...
Abstract The SAL cognitive architecture is a synthesis of two well-established constituents: ACT-R, hybrid symbolic-subsymbolic architecture, and Leabra, neural architecture. These component architectures have vastly different origins yet suggest surprisingly convergent view the brain, mind behaviour. Furthermore, both these are internally pluralistic, recognising that models at single level abstraction cannot capture required richness In this article, we offer brief principled defence...
This article presents a new research area called interactive task learning (ITL), in which an agent actively tries to learn not just how perform better but the actual definition of through natural interaction with human instructor while attempting task. The authors provide analysis desiderata for ITL systems, review related work, and discussion possible application areas systems.
Sensemaking is the active process of constructing a meaningful representation (i.e., making sense) some complex aspect world. In relation to intelligence analysis, sensemaking act finding and interpreting relevant facts amongst sea incoming reports, images, intelligence. We present cognitive model core information-foraging hypothesis-updating processes applied spatial probability estimation decision-making tasks. While was developed in hybrid symbolic-statistical architecture, its...
ABSTRACT In social interactions, decision makers are often unaware of their interdependence with others, precluding the realization shared long‐term benefits. an experiment, pairs participants played Iterated Prisoner's Dilemma under various conditions involving differing levels information. Each pair was assigned to one four conditions: “No‐Info” players saw own actions and outcomes, but were not told that they interacted another person; “Min‐Info” knew person still without seeing other's...
Abstract Recent research in cybersecurity has begun to develop active defense strategies using game‐theoretic optimization of the allocation limited defenses combined with deceptive signaling. These algorithms assume rational human behavior. However, behavior an online game designed simulate insider attack scenario shows that humans, playing role attackers, far more often than predicted under perfect rationality. We describe instance‐based learning cognitive model, built ACT‐R, accurately...