# Affiliated Faculty

Wojtek Chacholski , Professor, Mathematics

"I believe the shape of data does matter and to further our abilities to analyze data it is therefore essential that statistical methods are adjusted and developed to incorporate geometrical and topological information. How to transform such information into signatures suitable for statistical analysis and validation is what drives my interest. I am particularly interested in homology based invariants and new ways of using them for data analysis purposes."

David Eklund , PhD, Mathematics

"My research interest is computational algebraic geometry, in particular numerical methods and applications to science and engineering. One theme in my research is algorithms for invariants of algebraic varieties and I am currently exploring the possibility of using large data sets of points from algebraic varieties to compute such invariants."

Anders Forsgren , Professor, Optimization and Systems Theory

"My main research interest is nonlinear optimization, in particular method development and applications in radiation therapy. The aim is development of efficient methods. The aim is not immediately about handing of complex data, but I envisage that the techniques we develop might be of use also in this setting."

Ather Gattami, Senior Scientist, AI Expert, RISE SICS

"My main research interests are within big data analytics, deep learning(mathematical foundations, model compression, efficient algorithms, optimization), machine learning techniques for prediction and adaptive control of dynamical systems, re-inforcement learning, learning in (dynamic) games, signal reconstruction from sparse data, information theory and learning, with applications to NLPl, NLU, NLG, speech recognition, anomaly detection(fraud and anti-money laundering, diagnostics and predictive maintenance in manufacturing), and collaborative filtering (recommender engines)."

Patrick Henning , Assistant Professor, Numerical Analysis

"In my research I am concerned with the numerical treatment of partial differential equations that involve extreme data variations and multiple, largely different scales - so called multiscale problems. Here I focus on the development, analysis and implementation of new methods, particularly designed to treat such multiscale problems. Particular examples for applications that I am interested in are porous media flow, acoustic wave propagation in heterogenous media, negative refraction in photonic crystals and Bose-Einstein condensation."

Xiaoming Hu , Professor, Optimization and Systems Theory

"My main research interests are in complex and networked systems, active sensing and perception, control of multi-agent systems, nonlinear observer design, and mobile manipulation. A common key issue among these research areas is how to do modeling and control based on the large amount of data obtained from the various sensors that are usually noisy."

Anja Janssen , Associate Professor, Mathematical Statistics

"My research interest is extreme value theory and statistics, with a particular emphasis on extremal dependence of multivariate and time series data. My projects aim at a better understanding of the complex structures that we can find in extremal observations, by first providing adequate frameworks to formulate tail dependence and then use them to analyze data and theoretical model properties. I am especially interested in models for financial time series like GARCH and stochastic volatility models and their implications for the typical patterns which can be observed during extremal clusters."

Elias Jarlebring , Associate Professor, Numerical Analysis

"My research area is numerical linear algebra and matrix computations, i.e., the study of linear algebra for numerical computations. All computational problems in this field involve in some way huge amounts of data stored in matrices, vectors or tensor. The most common problems are linear systems of equations, eigenvalue problems, singular value problem, matrix equations, etc. These problems arise in most fields in science and technology, in particular in the topics of the lab, e.g., data analysis. In order to construct efficient and robust computational procedures for problems, we exploit the complex structures in the data (typically in the matrices)."

Johan Karlsson , Associate Professor, Optimization and Systems Theory

"My current research interests include: Analysis and identification of complex and stochastic systems with emphasis on reliability and robustness. Deformation models such as optimal mass transport for inverse problems with prior information and for modeling time-varying systems. Large scale optimization problems, moment problems and low complexity modeling. Applications in remote sensing, signal processing, and systems theory."

Anders Szepessy , Professor, Numerical Analysis

"My research interest is partial differential equations, numerical analysis and stochastic analysis. At the moment it is primarily directed towards analysis of classical and quantum mechanical particle systems, stochastic partial differential equations and inverse problems realted to optimal control. In all these areas there are research questions based on mathematics for complex data. For instance, reconstruction of empirical potentials from ab initio computations of particle systems, determination of distributions of stochastic parameters in differential equations and solving inverse problems with given data."

Richard Tsai , Professor, Numerical Analysis

"My research interests related to complex data include: the filtering and compression of data coming from exploration seismology, inverse scattering problems, segmentation and coarse graining algorithms for data on network, and analysis of point cloud data."

Bo Wahlberg , Professor, Automatic Control

"My research interests are in learning and control of dynamical system. Current research projects include matrix rank regularization techniques, sparse methods in Markov Decision Process problems, inverse filtering for Hidden Markov Models and identification of nonlinear stochastic dynamical systems."