Classification vs Regression Models for Creating prediction models for vaccination rates of incoming kindergarten students
Date
2021-04-28
Authors
Lee, Chen Yuan
Orellana, Erick
Zhu, Kening
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Vaccinations have been proven over the years to help treat some of the deadliest diseases as well as to prevent the further spread of those diseases. A major factor that can dictate the success or failure of a vaccine at treating the broader population is to reach heard immunity. To reach heard immunity, it is important to have a certain percentage of the population vaccinated so that they themselves will not be impacted by the disease and so that they will not spread throughout the community. Public health officials in conjunction with education officials understand the importance of immunizations and therefore set standards on vaccination rates and collect data to drive efforts to improve vaccination rates. The state of California is one of those states that has set standards for the vaccinations that kindergarten students should have before entering the school system and collect this information and make it public each year. There is an interest in using this information to make models that can be used to make prediction about vaccination levels at the county level. The most prominent approach taken is a geospatial approach of using a physical map to show vaccination rates. This is a useful visual, but it does not too much in the way of explaining why those vaccination rates are what they are based on certain factors. These geospatial models also do not provide a way to predict how changes in certain factors will impact the vaccination rates. In this study, variables that may impact vaccination rates are explored to generate a regression model and two classification models to understand if those models can be accurately used with the given predictor variables to gage potential changes to vaccination rates based on those given variables.
Description
Keywords
Classification, Decision trees, Regression models, Machine learning