Sequential Supervised Learning
Abstract:
We propose concepts for intelligent, adaptive cyber-physical systems that adaptively control information acquisition through the choice of sensor configurations to interact with the physical environment. Our focus is on a novel aspect of cyber-physical systems, intelligent sensing, to control information acquisition under sensing budget constraints in support of critical decision tasks. Our framework is to develop intelligent, adaptive controllers that trade the cost of alternative physical measurement configurations versus the value of the acquired information for improved decisions. These design tools are directed towards development of data-driven cyber-physical systems to account for scenarios where models for sensing, decision- making, and prediction are unavailable or poorly understood. Our recent work has investigated multi-stage decision systems to incorporate delays, costs, uncertainties and risks into decision- making for some limited but interesting settings that includes healthcare and homeland security applications. The tools developed will lead to understanding the inherent tradeoffs between risks, costs and delays when underlying models are poorly understood. From a conceptual perspective we develop new stochastic control methods for robustly optimizing costs when the underlying models are unknown.
Submitted by Venkatesh Saligrama
on
Abstract:
We propose concepts for intelligent, adaptive cyber-physical systems that adaptively control information acquisition through the choice of sensor configurations to interact with the physical environment. Our focus is on a novel aspect of cyber-physical systems, intelligent sensing, to control information acquisition under sensing budget constraints in support of critical decision tasks. Our framework is to develop intelligent, adaptive controllers that trade the cost of alternative physical measurement configurations versus the value of the acquired information for improved decisions. These design tools are directed towards development of data-driven cyber-physical systems to account for scenarios where models for sensing, decision- making, and prediction are unavailable or poorly understood. Our recent work has investigated multi-stage decision systems to incorporate delays, costs, uncertainties and risks into decision- making for some limited but interesting settings that includes healthcare and homeland security applications. The tools developed will lead to understanding the inherent tradeoffs between risks, costs and delays when underlying models are poorly understood. From a conceptual perspective we develop new stochastic control methods for robustly optimizing costs when the underlying models are unknown.