|
|
 |
| |
|
|
|
Efficiency of Multivariate Logistic Regression Analysis And Discriminant Analysis In Classification Of Rich Countries According To Human Development Index |
|
|
|
PP: 29-41 |
|
|
|
doi:10.18576/jsapl/050104
|
|
|
|
Author(s) |
|
|
|
Mohammed Mohammed Ahmed Al Mazah,
|
|
|
|
Abstract |
|
|
| In this paper, we used the multivariate logistic regression method to classify countries and to arrive at a linear model for distinguishing countries into rich and very rich, according to Human Development Index based on Human Development Report 2016. To illustrate the importance of the logistic regression model, it has been compared with the discriminating function, where they both classify the value to their correct society. We found through the results, that the method of logistic regression is more efficient than the discriminatory function in the classification of the countries under study. The logistic regression has been classified 92 countries out of 96 with a probability of 95.8% classification. The discriminating function classified 89 countries with a probability of 92.7%. , the comparison was relied on statistical criteria (apparent error rate, and apparent correct classification rate. The ordering of the influential indicators was significant on the classification of countries according to their relative importance in the case of the regression model (mean years of schooling, life expectancy at birth, and gross national income(GNI) per capita).In the case of the discriminant function (GNI per capita, maternal mortality, and life expectancy Birth). Based on this, we suggest that countries should be classified according to the United Nations Human Development Report and data processing using multi-response logistic regression models.
|
|
|
|
|
|
|
|
