Ulster University Logo

Ulster Institutional Repository

Stopping criterion impact on pure random search optimisation for intelligent device distribution

Biomedical Sciences Research Institute Computer Science Research Institute Environmental Sciences Research Institute Nanotechnology & Advanced Materials Research Institute

Poland, Michael, Nugent, CD, Wang, H and Chen, L (2010) Stopping criterion impact on pure random search optimisation for intelligent device distribution. In: International Conference on Intelligent Environments, Malaysia. IEEE. 6 pp. [Conference contribution]

Full text not available from this repository.

DOI: 10.1109/IE.2010.48

Abstract

The number of intelligent environmentimplementations such as smart homes is set to increasedramatically within the next 40 years. This is predicted usingforecasts of demographic data which indicates an expansion ofthe aged population. It has also been predicted that governmentswill struggle to meet the demand for resources such as sensortechnology due to costs. Optimisation of limited resourcesinvolves physically positioning devices to maximise pertinent-data gathering potential. Currently the most utilisedmethodology of distributing limited spatial detection sensors suchas pressure mats within smart homes is via ad-hoc deploymentsperformed by a human being. In this study idiosyncraticinhabitant spatial-frequency data was processed using a PureRandom Search (PRS) algorithm to uncover probabilistic futureregions of interest, alluding to optimal sensor distributions underresource constraint. With PRS a null hypothesis was stated:‘using lower iteration stopping criteria produce less optimalsensor distributions than when using higher iteration stoppingcriteria’. A student t-test between 1000 and 5000 iterations wasstatistically significant at 5% (p = 0.016852) whereby the nullhypothesis was rejected. Similar results were obtained betweenother iteration criteria. These data demonstrate that the iterationstopping criterion is not as critical as sensor size or number ofsensors; and that comparable results could be obtained whenlower stopping parameters are specified when using PRS.

Item Type:Conference contribution (Paper)
Faculties and Schools:Faculty of Computing & Engineering
Faculty of Computing & Engineering > School of Computing and Mathematics
Research Institutes and Groups:Computer Science Research Institute
Computer Science Research Institute > Artificial Intelligence and Applications
Computer Science Research Institute > Smart Environments
ID Code:15550
Deposited By:Professor Christopher Nugent
Deposited On:10 Sep 2010 14:43
Last Modified:10 Sep 2010 14:43

Repository Staff Only: item control page