Abstract
In this paper a HMM based method for analysing curves of amperometric biosensors is presented. In addition to our multi-HMM approach, we propose an enhanced single-HMM approach, which is used to detect the curve type, the end of the measurement and the correct measurement position. Besides, a problem-specific modified variant of the Baum-Welch algorithm is presented. The multi- and single-HMM approach yielded good accuracies and recognition rates. The modified Baum-Welch algorithm outperformed the standard Baum-Welch. © Springer-Verlag Berlin Heidelberg 2002.
Author supplied keywords
Cite
CITATION STYLE
Weitzenberg, J., Posch, S., & Rost, M. (2002). Analysis of amperometric biosensor curves using hidden-Markov-models. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2449 LNCS, pp. 182–189). Springer Verlag. https://doi.org/10.1007/3-540-45783-6_23
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
