番茄社区

Dr Casey Kearney

This course is compulsory on the MPA in Data Science for Public Policy. This course is not available as an outside option.

Students must have completed Pre-Sessional Coding and Mathematics Bootcamp (PP407).

This will ensure that students have basic fluency in Maths and Statistics along with Python and its main Data Science libraries. 

This course covers the theory and practice of the Data Science project lifecycle in Python for Public Policy, from problem definition and data sourcing/cleaning to exploration, visualization, and modelling. Emphasis will be placed on identifying problems that are suitable for different Data Science techniques and on good practices for managing data. Linear and logistic models and regularization techniques will be covered in the AT and Machine Learning, Clustering and introductory text analysis models will be left for the WT. Key concepts and ideas underlying modelling (bias vs. variance, types of error, training vs. test data) and data ethics and data science ethics will be illustrated and implemented with examples from healthcare, education, urban policy, international development, and other policy areas. By the end of the course, students will have a strong coding workflow and will be able to source and experiment with data for analysis and research, both individually and in a collaborative environment.

15 hours of lectures and 15 hours of seminars in the AT. 15 hours of lectures and 15 hours of seminars in the WT.

Students will be expected to produce weekly problem sets throughout the AT and WT.

1. James, Gareth, et al. An introduction to statistical learning: With applications in python. Springer Nature, 2023.
2. Chen, Jeffrey C., Edward A. Rubin, and Gary J. Cornwall. Data science for public policy. Springer, 2021.
3. Géron, Aurélien. Hands-on machine learning with Scikit-Learn, Ke