Predictive Analysis with Machine Learning 7.5 credits

This course is an introduction to machine learning and its application to make decisions in business and economics. The machine-learning methods are introduced by starting from a regression perspective and all methods covered are related to the standard regression analysis. The methods covered are regression analysis and classification (starting from the logistic regression-model). We include linear models, non-linear models and tree-based models. We also discuss regularization techniques such as lasso, ridge regression and elastic nets. Further, model selection techniques including information criteria and cross-validation are covered. We also cover bootstrapping methodology, which is a powerful tool for statistical inference. **Connection to Research and Practice** This course covers predictive modelling using machine-learning techniques. This is a fast-growing branch of statistics where analysis of big data is used for predictive and forecasting purposes. Most organizations today use big data for their decision making. The statistical methods introduced in this course enables government organizations, businesses etc. to use the data they collect in a strategic way to improve their operations. Further, they can be used in economic research directly to draw conclusions about unknown economic characteristics in the society.