Poverty Grade Evaluation Model Based on Multilevel Fuzzy System

Abstract

Poverty, a permanent problem in the society, is listed top-class global problem of social development by the United Nations. Grading the needy situation of impoverished families helps the government establish better policies, distribute resources more reasonably, and therefore provide aid more effectively. However, the traditional single-factor mode is not adequate, for poverty grade evaluation involves various factors of different weights, and some factors cannot be analyzed by classical algorithm. To overcome such problems, in this paper we establish a model applying the theories and methods of fuzzy mathematics and comprehensive evaluation. Based on fuzzy inference, we perform evaluations which are both qualitative and quantitative, and include exact and inexact factors. We determine the indexes of poverty grade according to maximum membership degree, and assign their weight using Analytic Hierarchy Process (AHP) -- in this way we quantify the qualitative problems. Finally, we verify our model with instances; the test result indicated that this technologically-advanced model provides a higher reliability to poverty grade evaluation, and is practically applicable.