Discovering Admissible Simultaneous Equations of Large Scale Systems

SSFisasystemtodiscoverthestructureofsimultaneous equations governing an objective process through experiments. SSF combined with another system SDS to discover a quantitative formula of a complete equation derives the quantitative model consisting of simultaneous equations re ectingtherstprinciplesunderlyingintheobjectiveprocess. ThepowerofSSFcomesfromthe use of the complete subset structure in a set of simultaneous equations which can be experimentally identied. The theoretical foundations of the structure identication and the algorithm of SSF are described, and its eciency and practicality are demonstrated and discussed with large scale working examples. This work is to promote the research of scientic discovery to a novel and promising direction, since the conventional equation discovery systems could not handle such a simultaneous equation process.

• Type: Conference
• Year: 1998
• Where: AAAI
• Authors: Takashi Washio, Hiroshi Motoda
• Keywords: AAAI 1998, Conventional Equation Discovery, Intelligent Agents, Simultaneous Equation, Simultaneous Equation Process