Selection of income indicators for Middle East country classification
| Autor: | Mariam Kalakech, Denis Hamad, Ali Kalakech |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
Computer science
Context (language use) 010103 numerical & computational mathematics Space (commercial competition) 01 natural sciences Outcome (probability) Electronic mail 010104 statistics & probability Matrix (mathematics) Discriminant Statistics 0101 mathematics Laplace operator Selection (genetic algorithm) |
| Zdroj: | 2016 Sixth International Conference on Digital Information Processing and Communications (ICDIPC). |
| DOI: | 10.1109/icdipc.2016.7470792 |
| Popis: | Laplacian score used to select the most relevant income (input) indicators for Middle East countries, has shown good classification performances of those countries, while reducing their input indicator space. In this paper, we propose a new way to calculate the similarity matrix used by this Laplacian score in order to perform the selection. This similarity matrix is calculated using selected outcome (output) indicators. Based on this matrix, a Laplacian score is attributed to each input indicator. These indicators are then ranked according to their scores and the most discriminant ones are selected. Results show the interest of the proposed approach for indicator selection to perform classification of those Middle East countries. They also reveal that the women participation is a critical dimension of the development process of a country. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|