Award: George Pólya
Year of Award: 1998
Publication Information: The College Mathematics Journal, Vol. 27, No. 1, (1997), pp. 2-19
Summary: A look at mnemonic techniques and neural networks through the construction of linear transformations by the accumulation of many small rank-one adjustments.
About the Author: (from The College Mathematics Journal, Vol. 27, No. 1, (1997))
Kevin Kirby’s Ph.D. is from Wayne State University (1988); in 1994 he joined the Mathematics and Computer Science Department at Northern Kentucky University. He has taught graduate courses on neural networks, but his main research interests are natural computability theory and the physics of information. His hobbies include piano, taekwondo, and symplectic geometry. One day in 1986 his advisor let him give a crazy guest lecture on neural networks in the Dirac notation; that lecture was the ancestor of this paper.