Algorithmic Systems

We aim at finding algorithmic solutions for science and society, and with this target in mind, our goal is to study the theory and practice of modelling, designing, and managing complex systems. The work has a strong basic research component that intersects artificial intelligence, machine learning, theoretical computer science, information theory and mathematical statistics, and the results of this methodological work are applied to both scientific and industrial applications.

Methodological approaches

Theoretical frameworks for probabilistic modeling. Development of computationally efficient, general-purpose methods for probabilistic modeling, focusing on issues related to model selection, parameter estimation and inference.
Real-time optimization. Approximate but fast inference methods combining deterministic and stochastic techniques.
Methods for data fusion. Probabilistic models for combining inputs originating from heterogeneous data sources.
Intelligent information access. Solutions to information retrieval tasks appearing in ubiquitous environments.
User modeling. Methods for personalization, profiling, segmentation and visualization.

Examples of current research projects

VISCI: Virtual Intelligent Space for Collaborative Innovation (Myllymäki, Floréen)

VISCI TOOLS: Tools for Virtual Collaborative Innovation (Myllymäki, Floréen)
AICA: Adaptive Interfaces for Consumer Applications (Floréen, Myllymäki)
LUCRE: Local and User-Created Services (Floréen)
WiSh: Widget Sharing (Floréen)
Online Optimisation and Production Planning (Orponen)
Algorithmics for Data Security (Orponen)
MCM: Methods for Constructing and Solving Large Constraint Models (Niemelä)

Programme management

Programme Director: Professor Petri Myllymäki
Programme Manager: Dr. Teemu Roos
Programme Management Group:
- Docent Patrik Floréen
- Professor Petri Myllymäki
- Professor Ilkka Niemelä
- Professor Pekka Orponen