Is this project an undergraduate, graduate, or faculty project?
Undergraduate
Project Type
individual
Authors' Class Standing
Daniel Oldham, Senior
Lead Presenter's Name
Daniel Oldham
Faculty Mentor Name
Mihhail Berezovski
Abstract
Data mining and statistical analysis software are increasingly becoming widespread in business fields to maximize company efficiency, and the technology itself has applications for a plethora of beneficial industry and real-world applications. Using clinical patient data from a foster care organization based in Gainesville, Florida, this research attempts to gain insights into childrens’ lives using R statistical analysis software and Orange Data Mining Suite. In total, 8 years’ worth of data totaling 250,000 observations and 60 variables regarding both the children and their parents was imported, cleaned, and leveraged for predictive analysis using these programs. The goal is to provide insights to help the organization identify the characteristics of the children and parents who are most at risk for undesirable outcomes. For example, although children’s cases are complex and unique, many of them end up “re-entering” back into the foster care system, even after being provided services, and this is what we attempted to predict and what the organization wants to eliminate as much as possible. Various insights to do this have been discovered, and as the research continues, more are being found. This research provides a glimpse into the realm of data mining for unique industrial purposes and sheds light on the diversity of state-of-the-art data mining and statistical programs’ capabilities.
Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, Collaborative, Climbing, or Ignite Grants) from the Office of Undergraduate Research?
Yes, Spark Grant
Data Mining to Benefit Foster Care Children and Parents
Data mining and statistical analysis software are increasingly becoming widespread in business fields to maximize company efficiency, and the technology itself has applications for a plethora of beneficial industry and real-world applications. Using clinical patient data from a foster care organization based in Gainesville, Florida, this research attempts to gain insights into childrens’ lives using R statistical analysis software and Orange Data Mining Suite. In total, 8 years’ worth of data totaling 250,000 observations and 60 variables regarding both the children and their parents was imported, cleaned, and leveraged for predictive analysis using these programs. The goal is to provide insights to help the organization identify the characteristics of the children and parents who are most at risk for undesirable outcomes. For example, although children’s cases are complex and unique, many of them end up “re-entering” back into the foster care system, even after being provided services, and this is what we attempted to predict and what the organization wants to eliminate as much as possible. Various insights to do this have been discovered, and as the research continues, more are being found. This research provides a glimpse into the realm of data mining for unique industrial purposes and sheds light on the diversity of state-of-the-art data mining and statistical programs’ capabilities.