ارسل ملاحظاتك

ارسل ملاحظاتك لنا







Bayesian Variable Selection for Single Index Model

المصدر: مجلة القادسية للعلوم الإدارية والاقتصادية
الناشر: جامعة القادسية - كلية الادارة والاقتصاد
المؤلف الرئيسي: Alshaybawee, Taha Hussain Ali (Author)
مؤلفين آخرين: Abed-Alghanemi, Rasha Majeed (Co-Author)
المجلد/العدد: مج25, ع2
محكمة: نعم
الدولة: العراق
التاريخ الميلادي: 2023
الصفحات: 237 - 244
ISSN: 1816-9171
رقم MD: 1399132
نوع المحتوى: بحوث ومقالات
اللغة: الإنجليزية
قواعد المعلومات: EcoLink
مواضيع:
كلمات المؤلف المفتاحية:
Nonparametric Bayesian Inference | Single Index | Semiparametric Model | MCMC | Scale Mixture of Uniform
رابط المحتوى:
صورة الغلاف QR قانون
حفظ في:
المستخلص: Nonparametric models with multivariate are suffer from the so called curse of dimensionality. Single index model one of the most attractive models use to overcome this problem, this model can reduce the dimensionality with holding a lot of flexibility of a nonparametric model, that what we need in each of econometrics and statistics. A Bayesian approach based on the asymmetric Laplace for estimation and variables selection was address by many researchers recently. In this paper, we proposed a new Bayesian lasso for single index model based on a scale mixture uniform. The Gaussian process prior have been considered to the nonparametric link function. Simulation examples (with different samples size and different modes) are considered, also real data for the proposed method show substantial improvements compare the other methods.

ISSN: 1816-9171

عناصر مشابهة