Harvard Catalyst Profiles

Harvard University

ScD

2001

Biostatistics

Dr. Erin (Kammann) Lake is Primary Faculty and Lecturer on Biostatistics, Co-Director of the Master of Science in Biostatistics (SM BIO) Programs, and the department lead on student development efforts.

Dr. Lake is trained in theoretical math and biostatistics, and her research centers around learning optimally and appropriately from data in order to improve public health and medicine. Current research includes:
- methods for classic statistical modeling, machine learning and AI
- 'alignment' in machine learning and AI (weak to strong, top-down interpretability, scalable oversight)
- pedagogy for biostatistics and data science
- application of statistical and machine learning methods in collaborative work across many disciplines at the Harvard Chan School and beyond.

Dr. Lake’s methodological contributions include adapting core machine learning methods to accommodate nuanced data scenarios (such as correlated outcomes and missing data), disease mapping with her development of the Geoadditive model and other robust smoothing methods, Lyme disease research via mathematical models around transmission and treatment, cluster-randomized trials involving pilot studies, and extensions of the Anderson-Gill model for repeated time-to-event data in drug/device development. She has also made important applied contributions in the areas of HIV and AIDS clinical trials, APOE genotype and Alzheimer disease, disability compensation and bone/joint injury, smoking cessation, diet quality and diabetes in rural/developing geographic regions, and mental health during COVID-19. She worked as a statistician at Massachusetts General Hospital and at the biotech Alkermes, where she helped lead the Phase III trial of Vivitrol, the now FDA-approved sustained-release formula of Naltrexone to treat opioid and alcohol dependency.

Dr. Lake has received numerous teaching awards, including Teaching Assistant of the Year at Harvard for two consecutive years as a doctoral student, and four straight years of Letters of High Rating as a faculty member in the Department of Biostatistics - most recently for her extensive curriculum development and teaching of the Fall BST 210 course (linear, generalized linear and survival models with extensions to machine learning and AI). She has been the recipient of the Distinguished Student Paper Award of the International Biometrics Society (ENAR), and the Excellence in Research Awardee of the Association for Women in Mathematics and Society for Industrial and Applied Mathematics. She also taught BST 254 Special Topics in Biostatistics (for doctoral students in biostatistics).

Dr. Lake is the recipient of the 2022 Dean's Award for Scientific Advancement (Pedagogy).

• statistical learning, machine learning, AI, statistical modeling, pedagogy, clinical trials