Abstract:The strong association of birth weight with infant mortality is the main focus of birth weight research, with the assumption that birth weight is a major determinant of infant survival. Studies on factors of low birth weight in Kenya have neglected the flexible approach of using smooth functions for some covariates in models. Such flexible approach reveals detailed relationship of covariates with the response. The study sought to investigate risk factors of low birth weight in Kenya by assuming a flexible approach for continuous covariates and geographical random effect. The study used semi parametric models to flexibly model the effects of selected covariates and spatial effects. Our spatial analysis is based on a flexible geo-additive model using the provinces as the geographic unit of analysis, which allows to separate smooth structured spatial effects from random effect. A Gaussian model for birth weight in grams and a binary logistic model for the binary outcome were fitted. Continuous covariates was modeled by the penalized (p) splines and spatial effects was smoothed by the two dimensional p-spline. The specific objectives of the study was to investigate factors of low birth weight in Kenya by taking into account the hierarchical nature of child birth weight factors using a Bayesian hierarchical model. The study used secondary data from Kenya 2014 demographic and health survey (KDHS) data. The study found that child birth order, mother weight and height are significant predictors of birth weight. Secondary education for mother, birth order categories 2-3 and 4-5, wealth index of richer family and mother height were significant predictors of child size at birth. The area associated with low birth weight was North Eastern and areas with increased risk to less than average size at birth were Central and Eastern. The study found support for the flexible modeling of some covariates that clearly have nonlinear influences. Nevertheless there was no strong support for inclusion of geographical spatial analysis. The spatial patterns and the maps generated could be used for targeting development efforts at a glance. These findings have important implications for targeting policy as well as the search for left-out variables that might account for these residual spatial patterns.
Investors, Policy makers, Governments etc. are all consumers of exchange rates data and thus exchange rate volatility is of great interest to them. Modeling foreign exchange (FOREX) rates is one of the most challenging research areas in modern time series prediction. Neural Network (NNs) are an alternative powerful data modeling tool that has ability to capture and represent complex input/output relationships. This study describes application of neural networks in modeling of the Kenyan currency (KES) exchange rates volatility against four foreign currencies namely; USA dollar (USD), European currency (EUR), Great Britain Pound (GBP) and Japanese Yen (JPY) foreign currencies. The general objective of the proposed study is to model the Kenyan exchange rate volatility and confirm applicability of neural network model in the forecasting of foreign exchange rates volatility. In our case the Multilayer Perceptron (MLP) neural networks with back-propagation learning algorithm will be employed. The specific objectives of the study is to build the neural network for the Kenyan exchange rate volatility and examine the properties of the network, finally to forecast the volatility against four other major currencies. The proposed study will use secondary data of the mean daily exchange rates between the major currencies quoted against the Kenyan shilling. The data will be acquired from the central bank of Kenya's (CBK) website collected for ten years of trading period between the years 2005 to 2017. The data will be analyzed using both descriptive and inferential statistics, with the aid of R's neuralnet package. A number of performance metrics will be employed to evaluate the model. Conclusion and recommendations will be made at the end of the study.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.