| Title |
Outlier detection in contingency tables / |
| Translation of Title |
Išskirčių aptikimas dažnių lentelėse. |
| Authors |
Jeriomova, Anastasija |
| Full Text |
|
| Pages |
30 |
| Keywords [eng] |
Raktiniai žodžiai : išskirtys, dažnių lentelės, grafiniai modeliai, jautrumas ir specifiškumas, Monte- Karlo modeliavimas Key words : outliers, contingency tables, graphical models, sensitivity and specificity, Monte-Carlo modeling |
| Abstract [eng] |
Outlier Detection in Contingency Tables In this work a new method for outlier detection in contingency tables is being examined, which is proposed in one of the most recent works in this field. The advantage of this method is that it is suitable not only for two-dimensional, but also for high-dimensional contingency tables. To our best knowledge, this type of method has not been proposed so far. The main goal of the thesis - by making use of Monte-Carlo modeling to investigate operational characteristics of the suggested method and to evaluate how these characteristics vary depending on table dimensions, the magnitude of deviation of outliers, the structural model used. The work provides with a brief overview of the method, a description of the modeling plan and the results obtained. |
| Dissertation Institution |
Vilniaus universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2021 |
