Decision Trees, Random Forests & A Case Study

Myles Gartland (http://www.rockhurst.edu/directory/faculty/myles-gartland/) will give a tutorial on using rpart (http://cran.r-project.org/web/packages/rpart/index.html) and randomForest (http://cran.r-project.org/web/packages/randomForest/index.html) in R. In addition, he'll also explain the concepts behind decisions trees and random forests. Decision trees are particularly useful due to their interpretability and ease of use. Random forests have gained quick acclaim for their ability to accurately predict outcomes in the face of high dimensionality and their for their ability to pick up on nonlinearities in data.

Myles is a professor and MBA director at Rockhurst University. He has a keen interest in data science and is an active practitioner of predictive modeling. He is Partner and Chief Analyst in Insightful Analytics, a predictive analytics and market research firm.

Matt Habiger (http://www.linkedin.com/profile/view?id=66978092&trk=nav_responsive_tab_profile_pic) will present a case study on predicting behavioral health readmissions (http://www.predictiveanalyticsworld.com/boston/2013/agenda.php#day1-510a). The case study will present an overview of how predictive analytics is done in an organization. His talk will cover how to define a predictive modeling problem, the process of collecting data in a business, building and deploying predictive models and measuring their impact to show value. The overarching goal of the talk is to highlight why data science is much more than just building a predictive model.