Mathematics & Computer Science
Investigation of Generalized Linear Mixed Models with Application to Salary Prediction
Document Type
Oral Presentation
Location
Indianapolis, IN
Start Date
13-4-2018 10:30 AM
End Date
13-4-2018 11:45 AM
Sponsor
Rasitha Jayasekare (Butler University)
Description
In linear models when the observations not independent due to different group structures or hierarchies that exist in data, the applicability of Generalized Linear Models (GLM) is violated. In such cases, Generalized Linear Mixed Models (GLMM) extend GLM to allow data that are not independent, by capturing both random and fixed effects of observations.
The goal of the project is to apply GLMM to develop a model that would predict salary as a binary response by identifying random effect variables within a data set of diverse individuals. The variables about individuals in the data set include things such as gender, race, sex, native country, occupation, and more. The validity of the model and accuracy of predictions are also discussed.
Investigation of Generalized Linear Mixed Models with Application to Salary Prediction
Indianapolis, IN
In linear models when the observations not independent due to different group structures or hierarchies that exist in data, the applicability of Generalized Linear Models (GLM) is violated. In such cases, Generalized Linear Mixed Models (GLMM) extend GLM to allow data that are not independent, by capturing both random and fixed effects of observations.
The goal of the project is to apply GLMM to develop a model that would predict salary as a binary response by identifying random effect variables within a data set of diverse individuals. The variables about individuals in the data set include things such as gender, race, sex, native country, occupation, and more. The validity of the model and accuracy of predictions are also discussed.