Redes neuronales para la estimación de la pobreza en el Ecuador

This paper proposes the use of neural networks tor the building of a proxy of poverty incidence in Ecuadorian urban areas, reproducing, with an important goodness to fit, the cffects ot macroeconomic policies in the short run, according to the data captured by the Surveys of Urban Employment, Undere...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: PÁEZ PÉREZ, PEDRO (author)
التنسيق: article
اللغة:spa
منشور في: 2020
الموضوعات:
الوصول للمادة أونلاين:https://estudioseconomicos.bce.fin.ec/index.php/RevistaCE/article/view/193
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
الوصف
الملخص:This paper proposes the use of neural networks tor the building of a proxy of poverty incidence in Ecuadorian urban areas, reproducing, with an important goodness to fit, the cffects ot macroeconomic policies in the short run, according to the data captured by the Surveys of Urban Employment, Underemployment and Unemployment. Although recognizing the importance of structural factors, the estimation confirms the testable predictions derived from the theoretical model. Tha<e poredictions say that exchange rate, wage and monctary policies, in that order, would have a decisive influence in poverty?s short run fluctuations. As a methodological contribution, this exercise shows the interesting possibilities that neural networks offer for estimation and modelization, even when, as in this case, we could have small sample problems.