Skip to main content
Browse by:
GROUP

Machine Learning Seminar

Event Image
Wednesday, January 25, 2017
3:30 am - 4:45 am
Jose Zubizarreta
Machine Learning Seminar

Machine Learning - Jose Zubizarreta (http://www.columbia.edu/~jz2313/)
New Matching Methods for Causal Inference and Impact Evaluation using Mathematical Programming
In observational studies of causal effects, matching methods are often used to approximate the ideal study that would be conducted if it were possible to do it by controlled experimentation. In this talk, I will discuss new matching methods based on mathematical programming that allow the investigator to overcome three limitations of standard matching approaches by: (i) directly obtaining flexible forms of covariate balance; (ii) producing self-weighting matched samples that are representative by design; and (iii) handling multiple treatment doses without resorting to a generalization of the propensity score. (iv) Unlike standard matching approaches, with these new matching methods typical estimators are root-n consistent under the usual conditions. I will illustrate the performance of these methods in real and simulated data sets. This is joint work with Magdalena Bennett, David Hirshberg and Juan Pablo Vielma.

Contact: Ariel Dawn