University of Minnesota
University Relations
http://www.umn.edu/urelate
612-624-6868
myU OneStop


Go to unit's home.

Home | Seminars and Symposia | Past seminars/symposia: Thursday, May 24, 2018

DTC Seminar Series

Computation in Nature and Lifelong Learning Machines

by

Hava Siegelmann
DARPA
Defense Advanded Research Projects Agency

Thursday, May 24, 2018
1:30 p.m. reception
2:00 p.m. seminar

401/402 Walter Library

Hava Siegelmann

Lifelong learning encompasses computational methods that allow systems to learn in runtime and apply previous learning to here-to-fore unexperienced situations without reprogramming or training. As this sort of computation is currently found almost exclusively in nature, lifelong learning looks to biology and neuroscience for its underlying principles and mechanisms.

Today’s most advanced AI, combining the preprogramming of Turing computation with limited learning via neural networks cannot fully account for biological learning. An alternative computational model suitable for lifelong learning is needed. We will discuss different forms of computation, and different computational mechanisms found in nature — including Super-Turing computation, stochastic and asynchronous communication, recurrent architectures, continual adaptivity, and interactive computation. While seemingly different, these varied computational attributes are, in fact, computationally equivalent, and in practice can provide, both singly and in combination, an underlying basis for lifelong learning.

We will discuss emerging thinking on practical implementation of biologically inspired concepts in AI and ML, underlying lifelong learning theory, as well as recent findings identifying a previously unrecognized brain connectome hierarchy that sheds light on varying levels of cognition and may underly abstract thought.

 

Dr. Siegelmann is a program manager at DARPA’s MTO office, developing programs to advance the fields of Neural Networks and Machine Learning. She is on leave from the University of Massachusetts where she serves as the director of the Biologically Inspired Neural and Dynamical Systems (BINDS) Laboratory, is a Professor of Computer Science and a Core Member of the Neuroscience and Behavior Program. She conducts interdisciplinary and cutting edge research in neural networks, machine learning, computational studies of the brain, intelligence and cognition, big data and industrial and biomedical applications.

Her research into neural processes has led to theoretical modeling and original algorithms capable of superior computation, and to more realistic, human-like intelligent systems. Siegelmann was named the 2016 Donald O. Hebb Award winner from the International Neural Network Society. Her Super-Turing theory of computation introduced a major variation in computational method and has become a sub-field of computation and the foundation of lifelong machine learning.

Super Turing also opens new ways to interpret cognitive mechanisms, as well as disease processes and their reversal. Her modeling of geometric neural clusters resulted in the highly utile and widely used Support Vector Clustering algorithm with Vladimir Vapnik and colleagues, specializing in the analysis of high-dimensional, big, complex data. Her neuroinformatics methods are used to identify overarching concepts of brain organization and function. A unifying theme underlying her research is the study of time and space dependent dynamical and complex systems. Her work is often interdisciplinary, combining complexity science, computational simulations, biological sciences and medicine — focusing on better and more complete modeling of human intelligence, and spanning medical, military and energy applications. Recent contributions include advanced human-machine interfaces that empower humans beyond regular capabilities, dynamical studies of the superchiasmatic nucleus and biological rhythm, and the study of brain structure that leads to abstract thoughts. Siegelmann was a recipient of the 2015 BRAIN initiative for her work on the energy constrained brain activation paradigm.

Dr. Siegelmann remains very active in supporting young researchers and encouraging minorities and women to enter and advance in STEM. She has designed and taught a variety of highly innovative interdisciplinary classes. She has years of experience consulting with industry, creating educational and international programs, fund raising, organization and management.