(P, p) retraining policies
(P, p) retraining policies
Skills that are practiced infrequently need to be retrained. A retraining policy is optimal if it minimizes the cost of keeping the probability that the skill is learned within two bounds. The (P, p) policy is to retrain only when the probability that the skill is learned has dropped just above the lower bound, so that this probability is brought up just below the upper bound. For minimum assumptions on the cost function, a set of two easy-to-check conditions involving the relearning and forgetting functions guarantees the optimality of the (P, p) policy. The conditions hold for power functions proposed in the psychology of learning and forgetting but not for exponential functions.
Dynamic programming, Instruction, Inventory management, Memory, Optimality
609-613
Katsikopoulos, Konstantinos V.
b97c23d9-8b24-4225-8da4-be7ac2a14fba
1 September 2007
Katsikopoulos, Konstantinos V.
b97c23d9-8b24-4225-8da4-be7ac2a14fba
Katsikopoulos, Konstantinos V.
(2007)
(P, p) retraining policies.
IEEE Transactions on Systems, Man, and Cybernetics Part A:Systems and Humans, 37 (5), .
(doi:10.1109/TSMCA.2007.902620).
Abstract
Skills that are practiced infrequently need to be retrained. A retraining policy is optimal if it minimizes the cost of keeping the probability that the skill is learned within two bounds. The (P, p) policy is to retrain only when the probability that the skill is learned has dropped just above the lower bound, so that this probability is brought up just below the upper bound. For minimum assumptions on the cost function, a set of two easy-to-check conditions involving the relearning and forgetting functions guarantees the optimality of the (P, p) policy. The conditions hold for power functions proposed in the psychology of learning and forgetting but not for exponential functions.
This record has no associated files available for download.
More information
e-pub ahead of print date: 20 August 2007
Published date: 1 September 2007
Keywords:
Dynamic programming, Instruction, Inventory management, Memory, Optimality
Identifiers
Local EPrints ID: 439265
URI: http://eprints.soton.ac.uk/id/eprint/439265
ISSN: 1083-4427
PURE UUID: 7771658b-5960-4fbb-a1a5-ffdf8c299f2e
Catalogue record
Date deposited: 07 Apr 2020 16:31
Last modified: 17 Mar 2024 03:44
Export record
Altmetrics
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics