15071

WiSe 23/24: Machine Learning for Political and Social Research

Ekoutiame Ahlonkor Ahlin

Kommentar

In a rapidly evolving world where data plays an increasingly pivotal role in understanding complex political and social dynamics, the course "Machine Learning for Political and Social Research" encompasses an extensive array of tools designed to unravel the complexities within data. These tools can be categorized into two main types: supervised and unsupervised. In a broader sense, supervised statistical learning revolves around constructing statistical models to predict or estimate outputs based on one or multiple inputs. On the other hand, unsupervised statistical learning deals with data containing inputs but lacking supervisory outputs. Despite this absence, the data still holds potential for us to uncover relationships and underlying structures. To exemplify the practical applications of machine learning, we will briefly explore real-world datasets examined within this course's context. The course will rely on the book "An Introduction to Statistical Learning" by Trevor Hastie, Robert Tibshirani, et al., and can be downloaded at https://www.statlearning.com/ . This course serves as a bridge between traditional statistical methods and cutting-edge machine-learning approaches. Through a balanced blend of theoretical concepts and practical applications, students will develop the skills necessary to navigate the intricacies of political and social data and derive meaningful insights. Course Highlights: • Fundamentals of Machine Learning o Introduction of machine learning o Assessing Models Accuracy • Linear Methods for Regression o Simple linear regression o Multiple linear regression • Classification Methods o Logistic regression o General Model for classification • Resample Methods o Cross-Validation o Bootstrap • Linear Model Selection and Regularization o Subset selection o Shrinkage methods • Support Vector Machine o Maximal Margin Classifier o Support Vector Classifiers • Unsupervised Learning Methods o Principal Component Analysis o Clustering Methods Schließen