Woody Bledsoe

from Wikipedia, the free encyclopedia
Woody Bledsoe

Woodrow Wilson "Woody" Bledsoe (born November 12, 1921 in Maysville , Oklahoma , † October 4, 1995 ) was an American mathematician and computer scientist who dealt with artificial intelligence , automatic proof and pattern recognition .

Origin and education

He was born into a large family on a small farm near Maysville. He only attended a small village school, which was interrupted when his father died in 1934 and he and his siblings had to take over the farm work completely. He ran away at the age of 16 and graduated from Calliham High School, South Texas, in 1939. In 1940 he moved back to his mother, who lived in Norman (Oklahoma) , and began his studies, which he financed as a dishwasher (with a daily 12-hour shift including Sundays). Soon afterwards he went to the US Army, completed his officer training and was involved as a pioneer officer in the 3rd US Army of George S. Patton von on the Rhine crossing to Germany (as a captain). He developed a method on site to quickly lower the heavy landing craft into the water. From 1945 he studied engineering and physics, then mathematics at the University of Utah , where he received his bachelor's degree in 1948. He then moved with a full scholarship to the University of California, Berkeley , where Alfred Tarski and John L. Kelley were among his teachers and he received his doctorate in 1953 with Anthony Perry Morse (Separative Measures for Topological Space).

Industrial activity

After his doctorate, he turned down courses at universities and went to the Sandia Corporation in Albuquerque (later Sandia National Laboratories ), where he worked as a mathematician and systems analyst on secret research in connection with the hydrogen bomb (he was in 1956 when testing a hydrogen bomb on the Eniwetok Atoll). At Sandia, too, he turned to pattern recognition, back then for the recognition of written or printed letters. With his Sandia colleague Iben Browning, he developed the Bledsoe-Browning N-tuple method published in 1959. The learning algorithm for letter recognition was adapted to a grid of N photocells, which were randomly connected in pairs and each represented 1 bit of information. In the following years he worked continuously on improving the algorithm. He left Sandia and with Iben Browning and Lloyd Lockingen (both from Sandia) founded their own company Panoramic Research Inc. (PRI) in Palo Alto in 1960 , which continued to work on these problems with orders from the Ministry of Defense and various secret services (but also, for example, for a Newspaper) worked.

During this time, Bledsoe had contacts with many pioneers of artificial intelligence in the USA ( Marvin Minsky , John McCarthy , Saul Amarel , Hans Joachim Bremermann ). In the early 1960s he experimented with Bremermann on genetic algorithms and evolutionary algorithms . During this time he also developed ideas with Bremermann about the influence of quantum mechanics on storage capacity and access time in computers. In 1964/65 he undertook secret work on facial recognition (with Helen Chan and Charles Bisson) for a secret service (comparison with photos stored in a large database, with a human operator still setting fixed points on the images at that time). After Bledsoe left PRI in 1966, the project was continued by Peter Hart at the Stanford Research Institute .

Professorship in Texas

Bledsoe wanted to return to academic research in 1965 and became a professor at the University of Texas at Austin . The initiative came from the university geneticists with whom he worked ( Wilson Stone ). At that time, gene locations on chromosomes were differentiated by the banding patterns, and pattern recognition algorithms were useful for research.

He was twice head of the mathematics faculty, first in 1967. At that time there was a faculty dispute between members of the then influential Moore method by Robert Lee Moore , who refused to use books and who did not want to buy any and who followed the Socratic method, and their opponents. Bledsoe supported the Moore method, which he himself had become acquainted with in Utah (and thus proved the existence of the Riemann integral for continuous functions, which led him to turn to mathematics). His second time as a director began in 1973. He stayed at the university until his retirement in 1994 (interrupted from his time as president of the AAAI until 1987).

When, in 1983, leading US computer companies decided to set up the MMC (Microelectronics and Computer Technology Corporation) consortium (under the leadership of former NSA director Bobby Ray Inman ) in response to the Japanese Fifth Generation Computer Project , which is based at the University of Texas at Austin was, Bledsoe was given the artificial intelligence department. There he sponsored Doug Lenat's Cyc project in particular .

Bledsoe himself continued to study automatic evidence with his students at the University of Texas, including developing the UT Prover program in the mid-1970s.

Robert S. Boyer is one of his PhD students .

Honors and memberships

In 1994 he received the Herbrand Award . In 1991 he received the IJCAI Distinguished Service Award and the Milestone Award from the American Mathematical Society.

In 1977 he was a board member of the International Joint Conference on Artificial Intelligence (IJCAI) and from 1978 to 1983 on its Board of Trustees. From 1983 he was President of the American Association for Artificial Intelligence (AAAI).

Private

He had been married to Virginia Bledsoe since 1944 and had three children. His wife was a Mormon from Salt Lake City, and Bledsoe also joined the Mormons, where he was twice elected bishop of his ward (serving as pastor but also caring for the social interests of the members), adviser to the president of his parish (Stake) and Stake became a patriarch . He was also an executive with the Boy Scouts of America .

Bledsoe died of ALS .

Fonts (selection)

Besides the writings cited in the footnotes:

  • The Model Method in Facial Recognition. Technical Report PRI 15, Panoramic Research, Inc., Palo Alto 1964
  • Some Results on Multicategory Pattern Recognition, J.ACM, Volume 13, 1966, pp. 304-316.
  • Splitting and Reduction Heuristics in Automatic Theorem Proving, Artif. Intell., Vol. 2, 1971, pp. 55-77.
  • A New Method for Proving Certain Presburger Formulas, Proc. IJCAI 1975, pp. 15-21.
  • Non-Resolution Theorem Proving, Artificial Intelligence, Volume 9, 1977, pp. 1-35.
  • with Michael Ballantyne :. Automatic Proofs and Theorems in Analysis Using Non-standard Techniques. Journal of the ACM, Volume 24, 1977, No. 3
  • with Michael Ballantyne: On Generating and Using Examples in Proof Discovery. Machine Intelligence, Volume 10, 1982
  • with Kenneth Kunen , Robert E. Shostak: Completeness Results for Inequality Provers, Artif. Intell., Vol. 27, 1985, pp. 255-288.
  • I Had a Dream: AAAI Presidential Address, August 19, 1985. AI Magazine, 1985

literature

  • Robert S. Boyer: Automated Reasoning: Essays in Honor of Woody Bledsoe. Kluwer Academic Publishers, 1991
  • Michael Ballantyne, Robert S. Boyer, Larry Hines: Woody Bledsoe: His Life and Legacy. AI Magazine, Volume 17, 1996, No. 1

Web links

Individual evidence

  1. ↑ Woody Bledsoe in the Mathematics Genealogy Project (English)Template: MathGenealogyProject / Maintenance / id used
  2. ^ Bledsoe, Browning, Pattern Recognition and Reading by Machine. In: Proceedings of the Eastern Joint Computer Conference 1959
  3. ^ Bledsoe, An Analysis of Genetic Populations. Technical Report, Panoramic Research Inc., Palo Alto, 1962
  4. ^ Bledsoe, The Evolutionary Method in Hill Climbing: Convergence Rates. Technical Report, Panoramic Research, Inc., Palo Alto, 1962
  5. ^ Bledsoe, A Basic Limitation on the Speed ​​of Digital Computers. IRE Transactions on Electronic Computers, Volume 10, No. 3, 1962
personal data
SURNAME Bledsoe, Woody
ALTERNATIVE NAMES Bledsoe, Woodrow Wilson (full name)
BRIEF DESCRIPTION American mathematician
DATE OF BIRTH November 12, 1921
PLACE OF BIRTH Maysville , Oklahoma
DATE OF DEATH 4th October 1995