Implementasi Regresi Logistik Bayesian Pada Indeks Pembangunan Manusia

The Human Development Index (HDI) is a measure of success in improving human welfare. The need of HDI as a tool for program design and evaluation of government policies in achieving development success targets. In increasing the HDI in a government, an effective and comprehensive program is needed as data in policy making, such as the use of the Bayesian logistic regression method. This research purpose to determine the factors influence HDI (Y) by involving the variables average length of school (X 1 ), expected length of school (X 2 ), per capita expenditure (X 3 ) with Markov Chains Monte Carlo – Gibbs Sampling computing. This method is an effective algorithmic computing technique which involves likelihood data and priors which are the posterior parameters of the model in HDI in East Java. The link function logistic regression models. Produces is a factor that has a significant effect on HDI, namely the average length of school (X 1 ). This factor indicates that the average number of years of schooling (X 1 ) has a positive value on HDI provided that other observed factors are constant. The results of this analysis conclude that the average length of schooling (X 1 ) is an important indicator in the success of increasing HDI in East Java.