Predicting Rare Diseases: Random Forests and Imbalanced Data by Nina Min '19, Wednesday, March 13

Predicting Rare Diseases: Random Forests and Imbalanced Data by Nina Min ’19, Wednesday, March 13, Statistics Colloquium today 1:10 – 1:50 pm, Stetson Court Classroom 105

Abstract:  Classification trees can be used to predict a response for future cases — like whether a new patient has a certain disease — based on previous observations. But building these classifiers using public datasets is challenging because the data are imbalanced: most people in the dataset don’t have the disease. We’ll approach this problem using random forests, an extension of classification trees, combined with a subsampling method for handling imbalanced data.