Final project for BDA course
Bayesian analysis over a dataset that regards the factors of incidence for serious suicide attempts in Shandong, China. The project is based on a study carried out in 2015 by a group of researchers.
Actually, we decided to carry out a Bayesian approach over the aforementioned dataset. Moreover, at the beginning, we chose to developed a frequentist approach, being that method covered during the last lectures in the course.
Therefore, adopting the bayesian approach we will first develop a multiple logistic regression model using all the variables. After that we will proceed with variable selection to determine which are the most influential factors, developing a second multiple logistic regression model using the selected variables.
Then, we will assess convergence analysis and efficency of the models, do posterior predictive checking and compare the models. To conclude we carry out a prediction on the age of the attempters and eventually answer our analysis problem.