Göran Falkman
Associate Professor of Computer Science
School of Informatics
Choose a course instance to see course syllabus and admission requirements.
There are no current course instances. If you have any questions, please contact the Course Coordinator or Study Counsellor.
The course focuses on the predictive analysis of time series and complex data. There are two challenges that are introduced: (i) dynamic behaviours that extend across time; (ii) multivariate systems that are interconnected, which can exhibit endogeneity. These go beyond typical cases explored in data mining and machine learning, where the unit of analysis is either a single entity or omitting temporal dynamics. The course focuses on introducing methods and methodologies for exploring, analysing and identifying structures in the data; models and modelling methodologies to deal with such data; evaluation frameworks and principles to ensure appropriate specification, selection and combination of predictions; and, theories and tools to interpret predictive models and predictions, so as to effectively validate and communicate them to users and stakeholders. In the course we explore various applications of predictive analysis, drawing from both societal and industrial challenges.