Quantitative Methods

The research performed by Prof. Dr. Lyudmila Grigoryeva is developing around the following two main areas of interest: (i) the theoretical foundations of machine learning for dynamic processes, particularly reservoir computing systems, and (ii) the modelling, classification and forecasting of time series with parametric and non-parametric machine learning models. She is especially interested in researching the accuracy of proposed approaches in specific applications, which range from the natural sciences to social sciences. In the context of economic sciences, the use of reliable and comprehensible learning models as a quantitative tool is of central importance as a result of the global social consequences of economic research. 

Research interests

• Reservoir Computing (RC), state-space systems
• Statistical learning and analysis of dynamic processes
• Machine learning (Recurrent Neural Networks, Deep Learning)
• Learning for dynamical systems and dynamical systems for learning
• Time series analysis, forecasting
• Financial econometrics

The research activities of Prof. Dr. Lyudmila Grigoryeva lead her to the interface of various disciplines and into a wide range of fields, extending from information technology to data science, economic sciences and finance.