Routine High-Return Human-Competitive Evolvable Hardware

This paper reviews the use of genetic programming as an automated invention machine for the synthesis of both the topology and sizing of analog electrical circuits. The paper focuses on the importance of the developmental representation in this process. The paper makes the point that genetic programming now routinely delivers high-return human-competitive machine intelligence. It also makes the point that genetic programming has delivered a progression of qualitatively more substantial results in synchrony with five approximately order-of-magnitude increases in the expenditure of computer time. The paper shows six examples where genetic programming has synthesized a circuit that duplicates the functionality or infringes a 21st century patented electrical circuit. Finally, the paper discusses how genetic programming can be enhanced in order to potentially enable it to deliver more complex industrial-strength results.

• Type: Conference
• Year: 2004
• Where: EH
• Authors: John R. Koza, Martin A. Keane, Matthew J. Streeter
• Keywords: EH 2004, Electrical Circuit, Genetic Programming, Hardware, Paper