Laboratory of Algorithmics
Theory and Practice of Algorithms
Algorithmics is the science and art of algorithms, i.e., of any type of procedure that computes something on hand. This comprises theory of computation (explaining which problems can and cannot be solved algorithmically), complexity theory (answering how good a problem can be solved), and the design and analysis of algorithms. The analysis of algorithms may be theoretical and/or experimental. Moreover, algorithmics deals with the role of randomness in the design and analysis of algorithms leading to probabilistic algorithms, quantum algorithms, and ultra-metric algorithms. Further important topics are parallel algorithms, distributed algorithms, and heuristics.
The field itself is very attractive and developing rapidly. This includes the development of new design techniques, new methods for the analysis of algorithms, and algorithmic engineering.
During the past decades the amount of data available in electronic form has grown enormously. For example, the WWW may have 30 billion pages, and it is estimated that more than 800MB of new information is produced and stored each year for each member of the human race. However, all these data constitute just an accumulation of knowledge fragments. But what is really needed is knowledge rather than fragments of knowledge. Processing these available data in traditional ways will lead nowhere, i.e., i.e., new technologies, new algorithms, new performance guarantees are sought for.
Another important aspect is information security. There are several conflicting goals, i.e., data should be available, data should be protected, data should be handled confidentially, and computer system threats should be avoided.
Our lab aims at contributing to all these topics of algorithms on the foundational and applied level. For example, we study algorithmic learning, algorithmic teaching, complexity theory, computability theory, data mining, QBF-solvers, parameter-free data mining, probabilistic algorithms, and very recently also ultra-metric algorithms. Moreover, we are interested in a variety of aspects of modern cryptology.