W5452 - Topics in Financial and Economic Data Science
This is an advanced seminar in data science which particularly covers the areas of modern statistical and econometric approaches as well as statistical and machine learning. Basic topics are on the application of suitable algorithms in those areas for modeling economic and financial data, especially for and forecasting economic and financial time series, based on known research results in the literature. For this purpose, new tools in modern areas of statistics and econometrics, such as local polynomial regression, P-Splines, quantile regression and functional data analysis, should be considered. Further new tools in recurrent neural networks, deep learning and reinforcement learning should be employed. Modelling and forecasting multivariate time series using proper adaptations of the above-mentioned approaches will also be studied. For high-level or research oriented seminar works more advanced topics, e.g. the extension of currently used methods in the literature for semiparametric modeling of long memory time series, deep learning of multivariate, functional or high-frequency financial and economic time series as well as Machine Learning algorithms for big financial and economic data can be offered.
Further Information can be found in the Modulhandbuch.