Machine Learning and the Starship - A Match Made in Heaven
P. Galea (2012), JBIS, 65, 278-282Refcode
: Machine learning, artificial intelligence, bayesian, neural network, social network analysis, interstellarAbstract:
The computer control system of an unmanned interstellar craft must deal with a variety of complex problems. For example, upon reaching the destination star, the computer may need to make assessments of the planets and other objects to prioritize the most `interesting', and assign appropriate probes to each. These decisions would normally be regarded as intelligent if they were made by humans. This paper looks at machine learning technologies currently deployed in non-aerospace contexts, such as book recommendation systems, dating websites and social network analysis, and investigates the ways in which they can be adapted for applications in the starship. This paper is a submission of the Project Icarus Study Group.
PDF file, 5 pages: £5.00