Cognitive Bias in Intelligent Systems
Systems Engineering and Systems Management Transformation
Dr. Laura Freeman
The recent advancement of artificial intelligence (AI) and machine learning (ML) have created a number of opportunities to employ these approaches to automated target recognition systems that go beyond the methods being used today. Advances in the application of AI and ML to tasks such as facial recognition, image and video tagging and metadata generation, and image fusion may present attractive opportunities to improve the performance of long-range intelligence, surveillance, and reconnaissance (ISR) systems. This research focuses on the concerns around inherent biases that come from training data, measurement approaches, and emergent properties of these learning systems that can make them vulnerable to false positive and false negative reports and that would make them inaccurate or easy to spoof. The Research Task ART-007 – Cognitive Bias in Intelligent Systems, which has been retitled to Performance Measures, Environments, Actuators, Sensors (PEAS) Framework for Test and Evaluation of Multi-Agent Systems of Autonomous Intelligent Agents to reflect the updated research agenda.
- Poster - RT 166: Formal Methods in Resilient Systems Design using a Flexible Contract Approach
- Poster - RT 182: Enterprise Systems-of-Systems Model for Digital Thread Enabled Acquisition
- Poster - Formal Methods in Resilient Systems Design using a Flexible Contract Approach
- Poster - Meshing Capability and Threat-based Science & Technology Resource Allocation