- author = "John R. Koza and Martin A. Keane and Matthew J. Streeter and William Mydlowec and Jessen Yu and Guido Lanza and David Fletcher",
- title = "Genetic Programming IV Video: Human-Competitive Machine Intelligence",
- year = "2003",
- publisher = "Kluwer Academic Publishers",
- keywords = "genetic algorithms, genetic programming",
-
video_url = "
https://youtu.be/ordtzUccs2s",
-
video_url = "http://www.human-competitive.org/sites/default/files/video/genetic_programming_4_video_human_competitive_machine_intelligence.mp4",
-
video_url = "http://gpbib.cs.ucl.ac.uk/koza_videos/genetic_programming_4_video_human_competitive_machine_intelligence.mp4",
- size = "42 minutes 32 seconds",
-
notes = "Accompanies \cite{koza:gp4}
The research reported in this book is a tour de force. For the first time since the idea was bandied about in the 1940s and the early 1950s, we have a set of examples of human-competitive automatic programming: John H. Holland.
This book demonstrates the every day solution of such holy grail problems as the automatic synthesis of analog circuits, the design of automatic controllers, and the automatic programming of computers. ... David E. Goldberg
It is amazing how this approach finds optimized solutions that are not obvious to the best human experts: Bernard Widrow, EE Stanford University
John Koza genetic programming approach to machine discovery can invent solutions to more complex specifications than any other I have seen: John McCarthy
2:23 GP now routinely delivers high-return human-competitive machine intelligence.
GP is an automated invention machine.
GP can automatically create a general solution to a problem in the form of a parameterised topology.
GP has delivered a progression of qualitatively more substantial results in synchrony with five approximately order-of-magnitude increased in the expenditure of computer time. (cf Moore law).
DVD menu (not in YouTube video) perhaps see \cite{koza:2009:gpt} or GP1 video https://youtu.be/tTMpKrKkYXo http://www.human-competitive.org/sites/default/files/video/genetic_programming_the_movie_part_1.mp4
9:36 The aim is to get machines to exhibit behaviour, which done by people, would be assumed to involve the use of intelligence: Authur Samuel
AI A to I ratio
Reuse!
11:00 de miniumus knowledge. High level goals. Routine.
13:00 radio antenna, gain, VSWR 14:00 Yagi 15:00 bridge design 16:00 metabolic pathways, chemical pathways, generation 225, 18:00 genetic networks 19:00 analogue electrical circuit, cf Gruau embryology. Modifiable components, including placement and routing. Campbell patent 22:00 square root function. 22:00 negative feedback: GP versus human intuition 24:00 cubic function 25:00 balun circuit 2001 patent 26:00 engineering around existing patents. Generating new inventions. 27:00 comprehensible or high performance? 28:00 free variable 29:00 generic plant controller. Economic savings. 31:00 PID tuning rules. 32:00 4 books on GP. Mooreware, Moore's law
36:00 reuse, self-organisation of hierarchies, ADF, passing parameters, architecture altering operations.
39:40 Invention is: Not deterministic. Not logical. Not knowledge-based. (GP non-greedy search, population-based, search in the space of computer programs, Biologically inspired.) Genetic diversity.
Koza videos \cite{koza:video} \cite{koza:video2} \cite{koza:video3} \cite{koza:video4} \cite{koza:2009:gpt}",
