|
Activity Number:
|
498
- Modern Machine Learning
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 6, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
|
Abstract #310989
|
|
|
Title:
|
Nonparametric Individual Treatment Effect Estimation for Survival Data with Random Forests
|
|
Author(s):
|
Denis Larocque* and Sami Tabib
|
|
Companies:
|
HEC Montreal and HEC Montreal
|
|
Keywords:
|
Random forest;
Individual treatment effect;
Survival data;
Censored data;
Tree-based method;
Heterogeneous treatment effect
|
|
Abstract:
|
Personalized medicine often relies on accurate estimation of a treatment effect for specific subjects. This estimation can be based on the subject's baseline covariates but additional complications arise for a time-to-event response subject to censoring. We propose new random forest methods for estimating the individual treatment effect with survival data. The treatment variable can be binary or continuous. The random forests are formed by individual trees built with splitting rules specifically designed to partition the data according to the individual treatment effect. For a new subject, the forest provides a set of similar subjects from the training data set that can be used to compute an estimation of the individual treatment effect with any adequate method. The merits of the proposed methods are investigated with a simulation study where it is compared to numerous competitors, including recent state-of-the-art methods. Examples of application with a colon cancer data and breast cancer data show that the proposed method can detect a treatment effect in a sub-population even when the overall effect is small or nonexistent.
|
Authors who are presenting talks have a * after their name.
