University of Twente Student Theses


Machine-learning algorithms in wave-confinement analysis

Schrijver, R. (2021) Machine-learning algorithms in wave-confinement analysis.

[img] PDF
Abstract:The application of clustering algorithms in the context of wave-confinement analysis is studied in this thesis. The algorithm is needed as a part of a supercell scaling method: a quantitative and algorithmic method of determining the confinement dimensionality of a state. Clustering validity indices are introduced as a means to gauge the clustering accuracy. We find that the accuracy of the clustering algorithm in our setting is measured best by the Davies-Bouldin index. Using the scaling method an alternative clustering approach is developed and its performance is judged. Ultimately, however, the k-means algorithm is deemed to perform better than the alternative one and is thus is selected as our clustering algorithm. We apply the combination of k-means clustering algorithm and the Davies-Bouldin index to analyse confinement properties of several configurations of the inverse woodpile photonic crystal structure. We conclude that this quantitative method defeats an existing method based on band structure analysis.
Item Type:Essay (Bachelor)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:30 exact sciences in general, 31 mathematics
Programme:Applied Mathematics BSc (56965)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page