OPCO 2065 - Data analysis with python - Advanced

The aim of this course is to deepen the students’ analytical skills and practical knowledge by diving into core

concepts of data analysis, statistics, and Python programming, thus enabling them to understand patterns in society

through data analysis. The course is especially focused on applications in the social sciences, including politics,

economics, health, and a wide range of social topics. By working with real-world datasets and using Python’s

advanced capabilities, students will learn to address key questions in society. Major topics in statistics and machine

learning will be covered, with leverage of libraries like numpy, scikit-learn, and pandas. The course has a special

focus on the analysis of textual data, with the use of state of the art machine learning models.

This course’s objectives are twofold: (I) the students will deepen their Python skills, thus including core concepts of

statistics and machine learning; (II) the students will create their own research design to investigate a real-world

dataset and put it to the test in a collective empirical project, thus also learning to effectively communicate findings.

This course is highly applied and deepens some statistical methods of data analysis, while the Intro class focuses on giving basic Python knowledge to address data analysis problems. Students who took the class “Introduction to data analysis with Python” during fall semester are welcome to enroll. Most generally, students are required to be familiar with Python (knowing how to import libraries, define functions, manipulate strings, lists, and dictionaries) and Jupyter notebook. Familiarity with pandas, scikit-learn, matplotlib and numpy is required. Students will have to take a small test at the beginning of the semester, to better evaluate their skills.

This courses is composed by strongly interactive, hands-on classes. Typically, one lesson will start with an explanation of the new concepts, both delving into theoretical concepts of data analysis, statistics, and machine learning, and showing their application. Then, the second part of the lesson will consist of exercises to do together, in a workshop modality. The slides and the sample code will be made available to the students on GitHub.

This course is evaluated on a pass/fail basis. Students are expected to complete the following

assessments in order to progress throughout the semester:

• After each lesson, students will receive homework exercises to complete. Homework can be

completed on a voluntary basis. Each homework sheet will be reviewed and corrected.

• To validate the course, students will work on a group project. Each group will choose a dataset

among the given ones and elaborate 2 to 3 research questions. The students will have to design a

pipeline to analyse the dataset according to their research question, visualize the results, and report a

qualitative and quantitative analysis. Both the paper and the code have to be submitted, and both will

be reviewed.

In Class Presence: 2 hours a week every 2 weeks / 12 hours a semester

• Repetition of concepts studied in class: 20 minutes every 2 weeks / 1 hour a semester
• Homework: 2h every 2 weeks / 10 hours a semester
• Research and Preparation for Group Work: 1 hour a week / 10 hours a semester
• Research and Writing for Individual Assessment: 20 minutes a week / 2 hours per semester