Targeted Learning

Causal Inference for Observational and Experimental Data

Targeted learning is a framework for causal and statistical inference methodology incorporating machine learning. This site contains research updates and resources. The book Targeted Learning: Causal Inference for Observational and Experimental Data, by Mark J. van der Laan and Sherri Rose, was published in 2011.

COMING 2017: Targeted Learning in Data Science by van der Laan and Rose, published by Springer.

Mark J. van der Laan is a Jiann-Ping Hsu/Karl E. Peace Professor of Biostatistics and Statistics at the University of California, Berkeley. His research concerns causal inference, prediction, adjusting for missing and censored data, and estimation based on high-dimensional observational and experimental biomedical and genomic data. He is the recipient of the 2005 COPSS Presidents' and Snedecor Awards, as well as the 2004 Spiegelman Award, and is a Founding Editor for the International Journal of Biostatistics.

Sherri Rose is an Associate Professor in the Department of Health Care Policy at Harvard Medical School. Her research interests include semiparametric estimation in causal inference and machine learning for prediction. She co-leads the Health Policy Data Science Lab at Harvard and is a 2016 AcademyHealth New Investigator. Dr. Rose serves on several editorial boards, including as associate editor for JASA-Theory & Methods and Biostatistics

---

When we started writing Targeted Learning, we went back and forth debating the level we were trying to target. Should we generate a textbook that was more like an epidemiology text and would be broadly accessible to a greater number of applied readers with less formal statistical training? Should we develop a purely theoretical text that would mostly be of interest to a certain subset of statisticians? Ultimately, we struck a level that is somewhere in between these two extremes. Since there is no other book on targeted learning, we could not escape the inclusion of statistical formalism. However, we also did not want to lose all accessibility for nontheoreticians.

This led to a book that begins with six chapters that should be generally readable by most applied researchers familiar with basic statistical concepts and traditional data analysis. The book progresses to more challenging topics and data structures, and follows a recognizable pattern via a road map for targeted learning and the general description of each targeted estimator. Thus, applied readers less interested in why it works and more interested in implementation can tease out those parts. Yet, mathematicians and theoretical statisticians will not get bored, as extensive rigor is included in many chapters, as well as a detailed appendix containing proofs and derivations.

Lastly, this book is unique in that it also contains wonderful contributions from multiple invited authors, yet it is not a traditional edited text. As the authors of Targeted Learning, we have spent significant time crafting and reworking each of the contributed chapters to have consistent style, content, format, and notation as well as a familiar road map. This yields a truly cohesive book that reads easily as one text.

Contributors

Antoine Chambaz (Paris Ouest Nanterre)
Victor De Gruttola (Harvard)
Iván Díaz (Google)
Bruce Fireman (Kaiser Permanente)
Susan Gruber (Harvard)
Alan Hubbard (Berkeley)
Nicholas Jewell (Berkeley)
Kelly Moore (The Gap)
Romain Neugebauer (Kaiser Permanente)
Maya Petersen (Berkeley)
Eric Polley (NCI)
Kristin Porter (MDRC)
Michael Rosenblum (Johns Hopkins)
Daniel Rubin (FDA)
Jasjeet Sekhon (Berkeley)
Michael Silverberg (Kaiser Permanente)
Richard Starmans (Utrecht)
Ori Stitelman (dstillery)
Catherine Tuglus (Amgen)
Hui Wang (VA Palo Alto)
Yue Wang (Novartis)
C. William Wester (Vanderbilt)
Wenjing Zheng (UCSF)