Radiopharmaceuticals, also called molecular imaging agents, if designed properly, can provide cross-sectional images of internal organs. These images can be used by physicians to diagnosis organ function. Modern molecular imaging agents accumulate within an organ by binding to a specific receptor. Consequently, the rate of accumulation is a function of the affinity and number of receptors within the organ.

The recent success of machine learning has been driven by advances in computer systems, and now it is time for a new era in which computer systems design is transformed through machine learning. This talk will focus on two of our recent works: Resource Allocation Optimization with Deep Reinforcement Learning (RL) and Dynamic Neural Networks with Sparsely Gated Mixture of Experts.

Edge computing near the sensor is preferred over the cloud due to privacy or latency concerns for a wide range of applications including robotics/drones, self-driving cars, smart Internet of Things, and portable/wearable electronics.  However, at the sensor there are often stringent constraints on energy consumption and cost in addition to throughput and accuracy requirements.

Current congestion of radio spectrum by licensed users and the increasing demand for new devices has created the need for alternative methods for usage of the radio spectrum. Licensed primary users can be inactive for a specific time, leaving their allocated frequency band unoccupied, which can be used by secondary users.  “White space” detection Cognitive radio (CR) devices are smart secondary users that are able to function properly when the primary users are inactive through dynamic spectrum access (DSA).

Abstract: Schizophrenia, Alzheimer’s disease, Parkinson’s disease and other degenerative disorders and dementias impose an enormous economic and psychosocial burden on society, communities, and families.