Monday, September 11, 2017

A00804 - Lotfi Zadeh, Father of Mathematical "Fuzzy Logic"

Photo
Lotfi Zadeh at his office at the University of California, Berkeley, in 1988.CreditCindy Manly-Fields, via University of California, Berkeley
Lotfi Zadeh, the computer scientist and electrical engineer whose theories of “fuzzy logic” rippled across academia and industry, influencing everything from linguistics, economics and medicine to air-conditioners, vacuum cleaners and rice cookers, died on Wednesday at his home in Berkeley, Calif. He was 96.
His son, Norman, confirmed the death.
Emerging from an academic paper Mr. Zadeh published in 1965 as a professor at the University of California, Berkeley, “fuzzy logic,” as he called it, was an ambitious effort to close the gap between mathematics and the intuitive way that humans talk, think and interact with the world.
If someone asks you to identify “a very tall man,” for instance, you can easily do so — even if you are not given a specific height. Similarly, you can balance a broom handle on your finger without calculating how far it can lean in one direction without toppling over.
Mr. Zadeh envisioned a mathematical framework that could mimic these human talents — that could deal with ambiguity and uncertainty in similar ways. Rather than creating strict boundaries for real world concepts, he made the boundaries “fuzzy.” Something was not in or out, for example. It sat somewhere on the continuum between in and out, and at any given moment a set of more complex rules defined inclusion.
“It was a bridge between theory and reality,” said Rudolf Seising, a professor at the University of Jena in Germany who specializes in fuzzy logic and worked alongside Professor Zadeh in his later years.
Continue reading the main story
In academic circles, Professor Zadeh’s work was controversial and sometimes ridiculed, in part because it challenged other forms of mathematics and in part because of his terminology. “Fuzzy logic” seemed to make fun of itself.
But the logic itself was not fuzzy, said Professor Timothy Ross of the University of New Mexico. It was a way of dealing with “fuzzy sets,” collections of information whose boundaries were vague or imprecise. Over the years it proved to be an enormously influential idea.
According to the website Google Scholar, Mr. Zadeh’s 1965 paper, titled “Fuzzy Sets,” has been cited by more than 90,000 scholarly works, and his mathematical concepts have provided practical new ways to build consumer electronics, trade stocks, forecast weather and more.
Lotfi Asker Zadeh was born on Feb. 4, 1921, in Baku, Azerbaijan, which was then a part of the Soviet Union. His father was a journalist, and his mother, born in Russia, was a doctor.
After the family moved across the Soviet border to Iran, Mr. Zadeh graduated with a science degree from the University of Tehran. During World War II he sold goods to the American Army, earning enough money to continue his education in the United States, his son said. He received a master’s degree in electrical engineering from the Massachusetts Institute of Technology in 1946 and a Ph.D. from Columbia in 1949.
Photo
Lotfi Zadeh in an undated photograph.CreditUniversity of California, Berkeley
As a professor at Columbia, working alongside John Ragazzini, Mr. Zadeh developed a mathematical method called z-transformations, which became a standard means of processing digital signals inside computers and other equipment. He moved to Berkeley in 1959.
Elijah Polak, a professor emeritus at Berkeley, recalled in an interview that Professor Zadeh’s theory of fuzzy sets emerged during their walks across the campus. Professor Zadeh began noticing that when people parked their cars, they would intuitively turn their wheels slightly to the left and then slightly to the right before pulling into a parking space.
“But how much is slightly?” he would ask.
Fuzzy sets began as an effort to use mathematics to define “slightly” — or “tall” or “fast” or “beautiful” or any other concept that has ambiguous boundaries.
Professor Zadeh originally envisioned fuzzy sets as simply a framework for harnessing language. But the idea expanded into other areas. It could provide a way for insurance companies to assess damage after an earthquake, for instance. Is the damage serious, moderate or minimal under company rules? Fuzzy sets could help.
“They opened up a whole new way of addressing problems where you don’t have precise data,” said Professor Ross, the author of a textbook on the practical uses of fuzzy logic.
The method could also help build machinery and electronics that gradually move from one state to another, like an automobile transmission, which shifts smoothly from first gear to second, or a thermostat, which flows just as smoothly from hot to cold. Hot and cold need not be precisely defined. They could exist on a continuum.
In the 1980s, Professor Zadeh’s ideas became popular among Japanese manufacturers, thanks to heavy investment from the government. Today the hype has faded, but fuzzy logic remains an active part of the mathematics that underpin the modern world.
In recognition of his work, Professor Zadeh received more than 50 engineering and academic awards. From 1963 to 1968 he was chairman of Berkeley’s electrical engineering department, helping to shift its focus toward computer science, a move that gave rise to one of the world’s top university computer science programs.
Professor Zadeh’s son is his only immediate survivor. His wife, Fania, died this year. A daughter, Stella, died in 2006. He was to be buried in Baku.
Professor Zadeh and others saw fuzzy logic as a tool for eventually building true artificial intelligence, and though many academics, including some Berkeley colleagues, questioned how effective these methods would be, he held firm.
“He always took criticism as a compliment,” said Stuart Russell, a Berkeley professor who worked next door to Mr. Zadeh for many years. “It meant that people were considering what he had to say.”
*******************************************************************

Lotfi Aliasker Zadeh[3] (/ˈzɑːd/AzerbaijaniLütfəli Rəhim oğlu Ələsgərzadə;[4] Persianلطفی علی‌عسکرزاده‎‎;[5] February 4, 1921 – September 6, 2017)[1][2] was a mathematician, computer scientist, electrical engineer, artificial intelligence researcher and professor emeritus[6] of computer science at the University of California, Berkeley.
He was best known for proposing fuzzy mathematics consisting of these fuzzy-related concepts: fuzzy sets,[7] fuzzy logic,[8] fuzzy algorithms,[9] fuzzy semantics,[10] fuzzy languages,[11] fuzzy control,[12] fuzzy systems,[13] fuzzy probabilities,[14] fuzzy events,[14] and fuzzy information.[15]
He was a founding member of the Eurasian Academy.[16]

Life and career[edit]

Zadeh was born in BakuAzerbaijan Soviet Socialist Republic,[17] as Lotfi Aliaskerzadeh,[18] to an Iranian Azerbaijanifather from Ardabil, Rahim Aleskerzade, who was a journalist on assignment from Iran, and a Russian Jewish mother, also an Iranian citizen,[19][20] Fanya Korenman, who was a pediatrician from Odessa.[21][22] The Soviet government at this time courted foreign correspondents, and the family lived well while in Baku.[23] Zadeh attended elementary school for three years there,[23] which he said "had a significant and long-lasting influence on my thinking and my way of looking at things."[24]
In 1931, when Zadeh was ten years old, his family moved to Tehran in Iran, his father's homeland. Zadeh was enrolled in Alborz College, which was a Presbyterian missionary school, where he was educated for the next eight years, and where he met his future wife, Fay.[23] Zadeh says that he was "deeply influenced" by the "extremely decent, fine, honest and helpful" missionaries from the United States who ran the college. "To me they represented the best that you could find in the United States – people from the Midwest with strong roots. They were really 'Good Samaritans' – willing to give of themselves for the benefit of others. So this kind of attitude influenced me deeply. It also instilled in me a deep desire to live in the United States."[24] During this time, Zadeh was awarded several patents.[23]
Despite being more fluent in Russian than in Persian, Zadeh sat for the national university exams and placed third in the entire country.[23] As a student, he ranked first in his class in his first two years. In 1942, he graduated from the University of Tehran with a degree in electrical engineering (Fanni), one of only three students in that field to graduate that year, due to the turmoil created by World War II, when the United States, Britain and the Soviet Union invaded Iran, whose ruler, Reza Shah, was pro-German. Over 30,000 American soldiers were based there, and Zadeh worked with his father, who did business with them as a contractor for hardware and building materials.[25]
In 1943, Zadeh decided to emigrate to the United States, and traveled to Philadelphia by way of Cairo after months of delay waiting for the proper papers or for the right ship to appear. He arrived in mid-1944, and entered the Massachusetts Institute of Technology (MIT) as a graduate student later that year.[25] While in the United States, he changed his name to Lotfi Asker Zadeh.[18]
He received an MS degree in electrical engineering from MIT in 1946, and then applied to Columbia University, as his parents had settled in New York City.[25] Columbia admitted him as a doctoral student, and offered him an instructorship as well.[25] He received his Ph.D. in electrical engineering from Columbia in 1949, and became an assistant professor the next year.[22][25]
Zadeh taught for ten years at Columbia, was promoted to Full Professor in 1957, and taught at the University of California, Berkeley from 1959 on. He published his seminal work on fuzzy sets in 1965, in which he detailed the mathematics of fuzzy set theory. In 1973 he proposed his theory of fuzzy logic.

Personal life and beliefs[edit]

Zadeh was called "quick to shrug off nationalism, insisting there are much deeper issues in life", and was quoted as saying in an interview: "The question really isn't whether I'm American, Russian, Iranian, Azerbaijani, or anything else. I've been shaped by all these people and cultures and I feel quite comfortable among all of them."[26] He noted in the same interview: "Obstinacy and tenacity. Not being afraid to get embroiled in controversy. That's very much a Turkish tradition. That's part of my character, too. I can be very stubborn. That's probably been beneficial for the development of Fuzzy Logic."[27] He described himself as "an American, mathematically oriented, electrical engineer of Iranian descent, born in Russia."[22]
Zadeh was married to Fay Zadeh and had two children, Stella Zadeh and Norman Zada.
Zadeh died on September 6, 2017, at the age of 96.[1][2] The previous month, the University of Tehran had released an erroneous report that Zadeh had died, but withdrew it several days later.[28]

Work[edit]

According to Google Scholar, as of September 2017, Zadeh's work has been cited about 180,000 times in scholarly works, with the 1965 "Fuzzy Sets" paper receiving about 90,000 citations.[29]

Fuzzy sets and systems[edit]

Zadeh, in his theory of fuzzy sets, proposed using a membership function (with a range covering the interval [0,1]) operating on the domain of all possible values. He proposed new operations for the calculus of logic and showed that fuzzy logic was a generalisation of classical and Boolean logic. He also proposed fuzzy numbersas a special case of fuzzy sets, as well as the corresponding rules for consistent mathematical operations (fuzzy arithmetic).[30]

Other contributions[edit]

Zadeh is also credited, along with John R. Ragazzini, in 1952, with having pioneered the development of the z-transform method in discrete time signal processing and analysis. These methods are now standard in digital signal processingdigital control, and other discrete-time systems used in industry and research. He was an editor of the International Journal of Computational Cognition.
Zadeh's most recent work included computing with words and perceptions. His recent papers include From Search Engines to Question-Answering Systems—The Role of Fuzzy Logic, Progress in Informatics, No. 1, 1-3, 2005; and Toward a Generalized Theory of Uncertainty (GTU)—An Outline, Information Sciences, Elsevier, Vol. 172, 1-40, 2005.

Selected publications[edit]

  • 1965. Fuzzy sets. Information and Control. 1965; 8: 338–353.
  • 1965. "Fuzzy sets and systems". In: Fox J, editor. System Theory. Brooklyn, NY: Polytechnic Press, 1965: 29–39.
  • 1972. "A fuzzy-set-theoretical interpretation of linguistic hedges". Journal of Cybernetics 1972; 2: 4–34.
  • 1973. "Outline of a new approach to the analysis of complex systems and decision processes". IEEE Trans. Systems, Man and Cybernetics, 1973; 3: 28–44.
  • 1974. "Fuzzy logic and its application to approximate reasoning". In: Information Processing 74, Proc. IFIP Congr. 1974 (3), pp. 591–594.
  • 1975. "Fuzzy logic and approximate reasoning". Synthese, 1975; 30: 407–428.
  • 1975. "Calculus of fuzzy restrictions". In: Zadeh LA, Fu KS, Tanaka K, Shimura M, editors. Fuzzy Sets and their Applications to Cognitive and Decision Processes. New York: Academic Press, 1975: 1–39.
  • 1975. "The concept of a linguistic variable and its application to approximate reasoning", I-III, Information Sciences 8 (1975) 199–251, 301–357; 9 (1976) 43–80.
  • 2002. "From computing with numbers to computing with words — from manipulation of measurements to manipulation of perceptions" in International Journal of Applied Math and Computer Science, pp. 307–324, vol. 12, no. 3, 2002.
  • 2012. Computing With Words. Principal Concepts and Ideas. Berlin: Springer, 2012.
A complete list of publications is on the website: http://www.cs.berkeley.edu/~zadeh/

Awards and honors[edit]

Awards received by Zadeh include, among many others:

No comments:

Post a Comment