Backpropagation Basics
May 1, 2022
Suki
May 1, 20221963
Ed Feigenbaum: was an artificial intelligence (AI) researcher addressed as the “father of expert systems.” He was co-winner of the 1994 Association of Computing Machinery Turing Award. His career highlights include co-editing with Julian Feldman Computers and Thinking, the first AI essay series. He is also remembered for his contributions to several significant programs, including medical systems, such as ACME, Mycin, SUMEX, and Dendral.(1)
ACME is one of the programs of the Medical Mortality Data System. It automates the underlying cause-of-death coding rules. ACME receives input as the cause-of-death codes (ICD) issued to each person (e.g., disease diagnosis, event, or injury) specified on cause-of-death certifications. ACME then adds WHO rules to the ICD codes and determines the underlying cause of death. ACME has been the de facto universal norm for automatically determining the root causes of death.(1)
MYCIN, named after the suffix of certain antimicrobial families, was an early backward chaining expert method. The program used machine learning to classify bacteria that cause serious infections, such as bacteremia and meningitis. Additionally, it advises antibiotics with weight-adjusted dose modifications for each patient.(2)
Dendral was a 1960s AI initiative that resulted in a machine science expert system. Its primary goal was to facilitate the creation and exploration of research hypotheses. It was applied on a selected primary science assignment: assist organic chemists in identifying unknown organic molecules by studying their mass spectra and contrasting them to previously characterized identified organic molecules utilizing chemistry information. Dendral is regarded as the first expert method because it automated organic chemists’ decision-making and problem-solving processes.(3)
Ivan Sutherland : wrote Sketchpad (a.k.a. Robot Draftsman) as his Ph.D. thesis in 1963. The thesis laid the groundwork for human-computer interaction (HCI). Sketchpad is widely considered the forerunner of computer-aided design (CAD) systems and a turning point in computer graphics advancement in general. Sketchpad is the basis of both the graphical user interface (GUI) and modern object-oriented programming. Ivan Sutherland’s legacy lies in the fact that he explained the usage of computer graphics for creative and technological uses and proposed new ways of human-computer interaction. For his contributions, he won the Turing Award in 1988 and, in 2012, received the Kyoto Prize.(3)
Ivan Sutherland : wrote Sketchpad (a.k.a. Robot Draftsman) as his Ph.D. thesis in 1963. The thesis laid the groundwork for human-computer interaction (HCI). Sketchpad is widely considered the forerunner of computer-aided design (CAD) systems and a turning point in computer graphics advancement in general. Sketchpad is the basis of both the graphical user interface (GUI) and modern object-oriented programming.
Ivan Sutherland’s legacy lies in the fact that he explained the usage of computer graphics for creative and technological uses and proposed new ways of human-computer interaction. For his contributions, he won the Turing Award in 1988 and, in 2012, received the Kyoto Prize.(3)
Leonard Uhr: was a computer vision, pattern analysis, machine learning, and cognitive science expert in the United States. He authored/edited eight books and almost 150 journal and conference articles in his fields. He pioneered various topics related to human neurophysiology and cognition and integrated AI algorithms and techniques with complex coping methods.
His thesis focused on developing AI systems inspired by his knowledge of how the human brain functions. His groundbreaking thesis, an article co-authored with Charles Vossler titled A Pattern Recognition Program That Generates, Evaluates, and Adjusts Its Own Operators, was published in 1956. It was reprinted in Computers and Thought — the collection of essays by the scientists who defined the AI field compiled by Feigenbaum and Feldman. (3)
Leonard Uhr: was a computer vision, pattern analysis, machine learning, and cognitive science expert in the United States. He authored/edited eight books and almost 150 journal and conference articles in his fields. He pioneered various topics related to human neurophysiology and cognition and integrated AI algorithms and techniques with complex coping methods.
His thesis focused on developing AI systems inspired by his knowledge of how the human brain functions. His groundbreaking thesis, an article co-authored with Charles Vossler titled A Pattern Recognition Program That Generates, Evaluates, and Adjusts Its Own Operators, was published in 1956. It was reprinted in Computers and Thought — the collection of essays by the scientists who defined the AI field compiled by Feigenbaum and Feldman. (3)
Thomas Evans: is another expert in the field of computation and AI. One of his most important contributions was ANALOGY, a piece of software written by Thomas Evans as part of his Ph.D. dissertation at the Massachusetts Institute of Technology. His thesis demonstrated that machines could solve analogy problems like those found on IQ tests.
ANALOGY was one of the first systems to attempt to mimic intelligence. In 1963 this program was a novel concept in computation and AI; nevertheless, it is now conventional. Evans wrote it to research the usage of “high-level” depictions of essential images. The ANALOGY software is worth analyzing since it demonstrates how various representations aid in accomplishing a goal. Its use of automated machines to solve “intelligent” problems (as opposed to, for example, gesture control or unconscious learning) elicited a distinct method now characteristic of AI.(3)