BIOSTAT 202: Opportunities and Challenges of Complex Biomedical Data: Introduction to the Science of "Big Data" (Summer 2020
Section outline
-
Lecture 1: Introduction to the course
What is data science/big data? What makes it different from non-big data? What big data can and cannot do. Phases of data science: getting data, merging and cleaning data, storing and accessing data, visualizing or telling stories with data, drawing conclusions from data.Faculty: Aaron Wolfe Scheffler
-
Live Interactive Discussion (Access restricted to registered students): Zoom URL. Brief formal review of lecture followed by question and answer discussion. Recorded lecture should be viewed prior to this session.
https://ucsf.zoom.us/j/99789000780?pwd=b3JVLzd2cWI5VTJNWkJEOEVDYnQ5UT09
-
Prerecorded Lecture 1 - module 1 Media Resource
-
Prerecorded Lecture 1 - module 2 Media Resource
-
Prerecorded Lecture 1 - module 3 Media Resource
-
Prerecorded Lecture 1 - module 4 Media Resource
-
Prerecorded Lecture 1 - module 5 Media Resource
-
Prerecorded Lecture 1 - module 6 Media Resource
-
-
Lab 1: Introduction to software
Live online session in which students have access to course faculty for questions on current or prior curriculum, assignments and software implementation
Faculty: Aaron Wolfe Scheffler
Location: https://ucsf.zoom.us/j/99789000780?pwd=b3JVLzd2cWI5VTJNWkJEOEVDYnQ5UT09-
Biostat 202 Assignment 1 Answer Key File
-
Biostat 202 Lab 1 - RECORDED Zoom Session URL
-
Lecture 2: Getting data from large databases
Public use data (NHANES, NIS). Electronic health records.
Faculty: Aaron Wolfe Scheffler
-
Biostat 202 Lecture 2 - RECORDED Zoom Session URL
-
Lecture 3: Managing data and data storage
Data management, cleaning, and storage.
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture 3 - RECORDED Zoom Session URL
-
-
Lab 2: Managing data and data storage
Live online session in which students have access to course faculty for questions on current or prior curriculum, assignments and software implementation
Faculty: Aaron Wolfe Scheffler
Location: Zoom URL-
Biostat 202 Lab 2 - RECORDED Zoom Session URL
-
Biostat 202 Assignment 2 Answer Key File
-
-
Lecture 4: Machine Learning
Introduction and supervised learning [regression]
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture 4 - RECORDED Zoom Session URL
-
-
Lecture 5: Machine Learning 2
Supervised learning [classification]
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture and Lab 5 - RECORDED Zoom Session 1/2 URL
-
Biostat 202 Lecture and Lab 5 - RECORDED Zoom Session 2/2 URL
-
-
Lab 3: Machine Learning
Live online session in which students have access to course faculty for questions on current or prior curriculum, assignments and software implementation
Faculty: Aaron Wolfe Scheffler
Location: Zoom URL-
Biostat 202 Assignment 3 Answer Key File
-
Lecture 6: Machine Learning 3
Supervised learning, cont. [classification]
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture 6 - module 2 URL
-
Biostat 202 Lecture 6 - module 3 URL
-
Biostat 202 Lecture 6 - module 4 URL
-
Biostat 202 Lecture 6 - RECORDED Zoom Session URL
-
-
Lecture 7: Machine Learning 4
Cross-validation, ensemble methods, and feature importance
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture 7 - module 1 URL
-
Biostat 202 Lecture 7 - module 2 URL
-
Biostat 202 Lecture 7 - module 3 URL
-
Biostat 202 Lecture and Lab 7- RECORDED Zoom Session URL
-
-
Lab 4: Machine Learning Competition
Live online session in which students have access to course faculty for questions on current or prior curriculum, assignments and software implementation
Faculty: Aaron Wolfe Scheffler
Location: Zoom URL-
Biostat 202 Lecture and Lab 7- RECORDED Zoom Session URL
-
Lecture 8: Unsupervised Learning
Unsupervised learning, clustering, and dimension reduction
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture 8 - module 1 URL
-
Biostat 202 Lecture 8 - module 2 URL
-
Biostat 202 Lecture 8 - module 3 URL
-
Biostat 202 Lecture 8 - RECORDED Zoom Session URL
-
-
Lecture 9: Causal Inference From Big Data
Issues of bias and how to minimize. Selection bias. Methods to minimize bias in observational studies.
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture 9 - module 1 URL
-
Biostat 202 Lecture 9 - module 2 URL
-
Biostat 202 Lecture 9 - module 3 URL
-
Biostat 202 Lecture and Lab 9 - RECORDED Zoom Session URL
-
-
Lab 5: Unsupervised Learning
Cover clustering and dimension reduction.
Faculty: Aaron Wolfe Scheffler
Location: Zoom URL-
Biostat 202 Assignment 5 Answer Key File
-
Lecture 10: Data visualization/storytelling
Graphical and tabular methods for displaying data to uncover/understand associations.
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Biostat 202 Lecture 10 - module 1 URL
-
Biostat 202 Lecture 10 - module 2 URL
-
Biostat 202 Lecture 10 - module 3 URL
-
Biostat 202 Lecture 10 - RECORDED Zoom Session URL
-
-
Lecture 11: Case studies in "Big Data"
Taking a step back, we will review what we learned and run through several case studies in "Big Data" and identify the possible ways to analyze the data.
Faculty: Aaron Wolfe Scheffler
-
Pre-recorded lecture - watch before live Zoom class session (Access restricted to registered students):
-
Big Data in Preterm Birth - Marina Sirota Case Study URL
-
Digital Health - Mark Pletcher Case Study URL
-
Biostat 202 Lecture and Lab 11 - RECORDED Zoom part 1 URL
-
Biostat 202 Lecture and Lab 11 - RECORDED Zoom part 2 URL
-
-
Lab 6: Data Visualization
Live online session in which students have access to course faculty for questions on current or prior curriculum, assignments and software implementation
Faculty: Aaron Wolfe Scheffler
Location: Zoom URL-
Biostat 202 Assignment 6 Answer Key File
-
Lecture 12: Project Office Hours
Faculty: Aaron Wolfe Scheffler
Ethics in Big DataEthical application of machine learning to big data is an important topic and I would like to provide some resources for you to archive and peruse as your own leisure. As clinical researchers, the approaches and considerations we take when analyzing data have implications on treatment and health outcomes. Even the most well intended applications of machine learning to big data can have unanticipated consequences that produce bias or even discrimination. Therefore, please keep these topics in mind as you develop your future research projects.Resources:Online courses:- Practical data ethics: an online short course on data ethics by https://ethics.fast.ai/. Very organized structure with recorded video lectures.
Books:- Cathy O'Neil. Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy. New York: Crown Publishers, 2016.
- Broussard, M. (2018). Artificial Unintelligence: How Computers Misunderstand the World. MIT Press.
Journal articles with discussion of bias in machine learning with clinical applications:
- Vyas DA, Eisenstein LG, Jones DS. Hidden in Plain Sight - Reconsidering the Use of Race Correction in Clinical Algorithms. N Engl J Med. 2020;383(9):874-882. doi:10.1056/NEJMms2004740
- Ziad Obermeyer and Sendhil Mullainathan. 2019. Dissecting Racial Bias in an Algorithm that Guides Health Decisions for 70 Million People. In Proceedings of the Conference on Fairness, Accountability, and Transparency (FAT* '19). Association for Computing Machinery, New York, NY, USA, 89. DOI:https://ucsf.idm.oclc.org/login?url=https://doi.org/10.1145/3287560.3287593

