Columbia Summer Research PREP: this team of scientist/engineer educators designed and taught an inquiry lab activity through ISEE’s Professional Development Program

Team Members: 

Lauren Anderson (Design Team Lead): Graduate Student, Biology, Flatiron Institute

Nick Young: Gradudate Student, Physics & Astronomy, Michigan State University

Rahul Jain: Gradudate Student, Physics & Astronomy, Michigan State University

Audience: 21 undergraduates

Activity name: Data-driven models; Decision Trees

Learning goal: 

Learners will develop the foundation of understanding supervised, data-driven models by building a decision tree to analyze and solve a predictive problem, and they will use accuracy as a metric to generate the best predictions. 

• A decision tree builds a prediction from multiple splits on features

• Test accuracy is a measure of the quality of the predictions

• There are optimal hyper-parameter values that maximize the test accuracy

This team was funded by: NSF AST#1743117 & Flatiron Institute