Nicholas Soures Email & Phone Number

Rochester, New York, United States

– Ph.D. Thesis: Lifelong Learning in Spiking Networks Through Neural Plasticity - Development of energy efficient, task agnostic lifelonglearning framework under realistic constraints– Derived novel regularization techniques to mitigate catastrophic interference and improve inter-task learning, outperforming similartechniques by up to 20% on image recognition tasks– Explored task-agnostic context-dependent processing through top-down attention– Co-developed custom digital and analog accelerators for on-device learning with metaplasticity supported by AFRL– Collaborated with University of Texas Health Science Center and Southwest Research Institute on epidemic modelling of COVID usinga combination of SEIR models and LSTM cells, providing insights and public access through interactive dashboard– Built machine learning framework for activity recognition via event-based cameras funded by NSA– Grant writing and reporting, helped secured near $2,000,000 in funding and led research efforts on several projects– Mentored graduate students in modelling, designing, and applying bio-inspired AI models and lifelong learning and was a TA for Brain-Inspired Computing– Helped develop Brain-Inspired Computing course including lectures, homework assignments, projects, and tutorials

Rochester, New York, United States

– Modeled hierarchical classification systems using transfer learning and autoencoders for real-world image datasets supported by NSA– Developed multi-layer neural networks for video-activity recognition which achieved state-of-the-art performance with 10x reductionin computation and memory compared to conventional networks supported by AFRL– Assessed the benefits of random CNNs for image processing to reduce energy and time requirements of CNN models on deployedsystems supported by AFRL– Conceptualized and modelled neuromorphic circuitry in both digital, analog, and photonic systems supported by AFRL and NSF– Analyzed of TrueNorth neuromorphic computing chip for image recognition with CNNs supported by NSA

Projects include:* Designing Neural Networks to forecast wind speed* Developing video games in Python* Designing a function generator able to produce square, triangle, and sine waves

TA for Physics 132

Doctor Of Philosophy (Ph.D.), Computer Engineering, 3.7

Activities and Societies: Research focus: Unsupervised/Supervised Learning in Spiking Neural Networks Semantic Knowledge Representation.

Master’S Degree, Computer Engineering, 3.7

Activities and Societies: My masters thesis was on "Unsupervised Deep Liquid State Machines for On-Device Intelligence" which investigated.

Bachelor Of Science (Bs), Physics

Activities and Societies: Club Hockey HackBU Physics Journal ClubRelevant Courses: Electronics 1, Mathematical Methods in Electrical.