Conference

Keynote title

Active Networked Intelligence and Hypothesis-driven Inference and Learning

Abstract

Beyond existing applications of AI, there is a critical need for artificial intelligence models and methodology to accurately and proactively interpret the physical world and assist us to monitor our ever-complex and large-scale industrial footprint on our planet. To achieve this, we need to simultaneously acquire data across large physical spaces AND actively interpret, in time, the diverse range and resolution of sensory inputs that make up the physical world. In this work, I will first discusses how this can be achieved by an integrated approach to connected, embodied and generative AI. Furthermore, I will discuss how advances in future networks design, including ubiquitous connectivity, integrated sensing and communication, and AI-enabled network management bring this to reality at scale.

Bios

Tara Javidi is an Iranian electrical engineer and computer scientist who studies networked information, stochastic control, machine learning, hypothesis testing, network optimization, and network routing, among other topics. She is a professor of electrical and computer engineering at the University of California, San Diego, where she co-directs the Center for Machine-Integrated Computing and Security with Farinaz Koushanfar.

Javidi was named a Fellow of the IEEE in 2021, "for contributions to stochastic resource allocation and active hypothesis testing". In 2020, the University of Michigan Department of Electrical and Computer Engineering recognized her with their Distinguished Educator Award, for her "significant and lasting impact in education".