| Autor: |
Eugenia Stoimenova, Nikolay Nikolov |
| Rok vydání: |
2021 |
| Předmět: |
|
| Zdroj: |
Advanced Computing in Industrial Mathematics ISBN: 9783030716158 |
| DOI: |
10.1007/978-3-030-71616-5_27 |
| Popis: |
In this paper, we consider a cluster analysis for complete rankings of N items that aims to identify typical groups of rank choices. The “K-means” procedure based on Lee distance is studied in details and several asymptotical results for large values of N are derived. An algorithm for approximating the normalizing constant in the clustering procedure is proposed by using some properties of Lee distance. In order to compare the clustering method based on Lee distance to those based on other distances on permutations, we apply the presented procedure to a data set obtained from the results of the American Psychological Association presidential election. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
