Beyond Tech - Machine Learning in Science & Policy - Professor David Dunson, Duke
Alan Turing Institute via YouTube
Overview
This course aims to highlight the limitations of applying off-the-shelf machine learning algorithms from the tech industry to fields like science and policy making. The learning outcomes include understanding the crucial issues of selection bias, uncertainty quantification, and limited training data, and developing targeted methods to address these challenges. Participants will learn skills such as removing sensitive variables for fair predictive algorithms and creating interpretable predictive models based on complex observations. The teaching method involves a lecture format with a focus on real-world applications in areas like criminal justice, neuroscience, policy, and health decision making. This course is intended for individuals interested in the intersection of statistics, mathematics, and computer science, particularly those working in fields such as epidemiology, environmental health, neurosciences, genetics, and policy making.
Syllabus
Intro
What is machine learning?
Some examples of labeled data
Mimicking automating humans?
Self-driving cars
What creates critical problems for deep learning?
Dealing with the data deluge in science
ML in policy & automated decision making
A geometric solution
Application to creativity
Taught by
Alan Turing Institute