TY - GEN
T1 - On ensemble techniques for AIXI approximation
AU - Veness, Joel
AU - Sunehag, Peter
AU - Hutter, Marcus
PY - 2012
Y1 - 2012
N2 - One of the key challenges in AIXI approximation is model class approximation - i.e. how to meaningfully approximate Solomonoff Induction without requiring an infeasible amount of computation? This paper advocates a bottom-up approach to this problem, by describing a number of principled ensemble techniques for approximate AIXI agents. Each technique works by efficiently combining a set of existing environment models into a single, more powerful model. These techniques have the potential to play an important role in future AIXI approximations.
AB - One of the key challenges in AIXI approximation is model class approximation - i.e. how to meaningfully approximate Solomonoff Induction without requiring an infeasible amount of computation? This paper advocates a bottom-up approach to this problem, by describing a number of principled ensemble techniques for approximate AIXI agents. Each technique works by efficiently combining a set of existing environment models into a single, more powerful model. These techniques have the potential to play an important role in future AIXI approximations.
UR - http://www.scopus.com/inward/record.url?scp=84871371276&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-35506-6_35
DO - 10.1007/978-3-642-35506-6_35
M3 - Conference contribution
SN - 9783642355059
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 341
EP - 351
BT - Artificial General Intelligence - 5th International Conference, AGI 2012, Proceedings
T2 - 5th International Conference on Artificial General Intelligence, AGI 2012
Y2 - 8 December 2012 through 11 December 2012
ER -