Diagnosing Old MI by Searching for a Linear Boundary in the Space of Principal Components

Mark Donnelly, Christopher Nugent, Dewar Finlay, Niall Rooney, Norman Black

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)
Original languageEnglish
Pages (from-to)476-483
JournalIEEE Transactions on Information Technology in BioMedicine
Issue number3
Publication statusPublished (in print/issue) - 1 Jul 2006

Bibliographical note

Other Details
This paper describes a study to improve the performance of perceptron-based classifiers for cardiac data analysis through the application of principal component analysis with novel weight initialisation. This work supports ongoing research to improve means of pervasive computing through reduction of computationally intensive analysis of high dimensional cardiac data. This paper is an extended version of one presented at IEEE CBMS 2005 and invited for consideration for the Special Issue on Mining Biomedical Data. The work, coupled with multi-lead electrocardiogram analysis, forms the basis of a forthcoming submission (November 2007) to the New Dynamics of Ageing Research programme.

Cite this