Statistics for Lunch Training Series: Decision Trees, Survival Trees, and Random Forest
To Know
Where:
Building 549 Executive Board Room
Organizer:
ABCS group
Presented By:
Brian Luke (Advanced Biomedical Computational Science, ABCS)
About this Class
This presentation describes the basic mechanics of constructing a decision tree and interpreting the results. Decision trees will be used to determine factors associated with 2-year survival for individuals with Stage IV lung cancer and whether Kyphosis (a type of deformation) occurs in children after spinal cord surgery. A relative hazard can also be used within a decision tree to construct a survival tree. Finally, random forest, which constructs a large number of sub-optimal decision trees, will be discussed. This presentation will be very light on mathematics, and no prior knowledge is required.