Random forest is in many fields of research a common method for data driven predictions. Within economics and prediction of poverty, random forest is rarely used. Comparing out-of-sample predictions in surveys for same year in six countries shows that random forest is often more accurate than current common practice (multiple imputations with variables selected by stepwise and Lasso), suggesting that this method could contribute to better poverty predictions. However, none of the methods consistently provides accurate predictions of poverty over time, highlighting that technical model fitting by any method within a single year is not always, by itself, sufficient for accurate predictions of poverty over time.
تفاصيل
-
المؤلف
-
تاريخ الوثيقة
2016/03/18
-
نوع الوثيقة
ورقة عمل خاصة ببحوث السياسات
-
رقم التقرير
WPS7612
-
مجلد رقم
1
-
عدد المجلدات
1
-
البلد
-
المنطقة
-
تاريخ الإفصاح
2016/03/18
-
حالة الافصاح
Disclosed
-
اسم الوثيقة
Is random forest a superior methodology for predicting poverty ? an empirical assessment
-
كلمة أساسية
small area estimation;development research group;labor force survey;impact of migration;impact on poverty;department of economics;loss function;rural area;simple average;consumption;selection method;estimation method;machine learning;consumption datum;total sample;confidence interval;regression equation;linear regression;development policy;pattern recognition;agricultural growth;occupational mobility;national poverty;spatial poverty;model prediction;poor household;public policy;data mining;open access;poverty targeting;model fitting;consumption proxy;consumption poverty;
- انظر المزيد
تنزيل الملفات
تقرير كامل
نسخة رسمية من الوثيقة (قد تضم توقيعات، الخ)
- نسخة رسمية (PDF)
- TXT*
- Total Downloads** :
- Download Stats