UC Berkeley Center for Targeted Learning

UC Berkeley Center for Targeted Learning

The UC Berkeley Center for Targeted Learning (CTL) harnesses the power of big data and statistical machine learning to improve health. CTL leverages unique developments in statistical machine learning, methodology pioneered by experts in the UC Berkeley Biostatistics Group, towards adaptation of these methods in research and applications.

The Targeted Learning approach provides a template to construct optimal machine learning algorithms for answering any type of (often causal) question about any type of observed data system, while providing formal statistical inference. Thus, the potential sphere of relevant applications is infinite, covering an enormous range of randomized trials, sequentially adaptive randomized trials, and complex longitudinal observational studies. Indeed, there exists strong interest in adaptation of these methods, including from the FDA, industry and researchers.

However, the uptake of specific applications of the targeted learning methods (beyond prediction) has been slowed due to educational and computational barriers. Our efforts to attack this challenge have taken three overlapping paths: 1) development of new software tools (in R) that can be both applied by end-users to analyze their data with relative ease (see https://github.com/tlverse), 2) to apply these methods in collaborative research, by working closely with groups to assist in analyzing their data, and 3) training for translation of these methods into practice (see for instance https://tlverse.org/acic2019-workshop/).

Current global health collaborations include funded projects with the Gates Foundation and the International Inter-American Development Bank (IADB).

Data Science
Causal Inference and Machine Learning

Alan Hubbard

Dr. Alan Hubbard is Professor of Biostatistics, Head of the Division of Biostatistics at UC Berkeley, and Head of data analytics core at UC Berkeley SuperFund. His current research interests include causal inference, variable importance analysis, statistical machine learning, estimation of and inference for data-adaptive statistical target parameters, and targeted minimum loss-based estimation. Research in his group is generally motivated by applications to problems in computational biology, epidemiology, and precision medicine.

Contact Information