Targeted Learning in Data Science

This textbook for Masters and PhD graduate students in biostatistics, statistics, data science, and epidemiology deals with the practical challenges that come with big, complex, and dynamic data while maintaining a strong theoretical foundation. It presents a scientific roadmap to translate real-world data science applications into formal statistical estimation problems by using the general template of targeted maximum likelihood estimators to construct targeted machine learning algorithms that incorporate state of the art applications to estimate quantities of interest, while still providing valid inference. Targeted learning methods within data science are a critical component for answering complex statistical questions in the modern age. The techniques can answer complex questions including optimal rules for assigning treatment based on longitudinal data involving time-dependent confounding and censoring as well as other estimands in dependent data structures, such as networks. Standard methods and software tools are not currently equipped for these challenges; however, targeted learning is tailored for these problems found in precision medicine, big data, and data science. Included in Targeted Learning in Data Science are demonstrations with software packages and real data sets that present a case that targeted learning is crucial for the next generation of statisticians and data scientists.