News from our scientific working group

We congratulate Dr. Atif Raza to his successful Ph.D. thesis defense

On the 13th of April 2021, our group member Atif Raza successfully defended his Ph.D. thesis, titled Metaheuristics for Pattern Mining in Big Sequence Data.

Interested readers can find the thesis here.

An overview of the thesis is given below:
An ever-growing list of human endeavors in a variety of domains results in the generation of time-series data, i.e., data that are time-resolved and measured in equidistant time intervals. The continued developments in sensor and storage technology and the availability of database systems specifically designed for time-series data have also made it possible to record an exorbitant amount of such data. The vast yet readily available data places ever-increasing demands on data mining methods for fast and efficient knowledge discovery, which establishes the need for exceedingly fast algorithms.
The data mining research community has been actively investigating various avenues to develop algorithms for time series classification. Most research has focused on optimizing accuracy or error rate, although runtime performance and broad applicability are as important in practice. The result is a plethora of algorithms that have quadratic or higher computational complexities. Consequently, the algorithms have little to no use for deployment on a large scale.
This thesis addresses the complexity issue by introducing several time-series classification methods based on metaheuristics and randomized approaches to improve the state-of-the-art in time-series mining. We introduce three subsequence-based time series classification algorithms and an approximate distance measure for time series data. One subsequences-based time series classifier explicitly employs random sampling for subsequence discovery. The other two subsequences-based classifiers employ discretized time series data coupled with (i) a linear time and space string mining algorithm for extracting frequent patterns and (ii) a novel pattern sampling approach for discovering frequent patterns. The frequent patterns are translated back to subsequences for model induction. Both of these algorithms are up to two orders of magnitude faster than previous state-of-the-art algorithms. An extensive set of experiments establishes the effectiveness and classification accuracy of these methods against established and recently proposed methods.
Posted on | Posted in News

Activities of the Data Mining Group at ECML/PKDD 2019

The Data Mining Group actively participated in a number of activities at the ECML/PKDD 2019 conference.

The group members presented research papers, posters and also organized two workshops. The workshop programs are available at the following links.

  • DeCoDeML (Deep Continuous-Discrete Machine Learning)
  • AIMLAI & XKDD (Advances in Interpretable Machine Learning and Artificial Intelligence & eXplainable Knowledge Discovery in Data mining)

The research paper "Pairwise Learning to Rank by Neural Networks Revisited: Reconstruction, Theoretical Analysis and Practical Performance: Marius Köppel*, Alexander Segner*, Martin Wagener*, Lukas Pensel*, Andreas Karwath^, Stefan Kramer*" was presented at the Ranking session.

The following papers were presented at the DeCoDeML and SoGood workshops.

DeCoDeML Spotlight Talk 1: Zahra Ahmadi, Sina Malakouti and Stefan Kramer. Deep Tree Networks: A New Symbolic Deep Architecture

DeCoDeML Spotlight Talk 3: Sophie Burkhardt, Nicolas Wagner, Johannes Fürnkranz and Stefan Kramer. Extracting Rules with Adaptable Complexity from Neural Networks using K-Term DNF Optimization

DeCoDeML Spotlight Talk 4: Nicolas Wagner, Sophie Burkhardt, Stefan Kramer. A Deep Convolutional DNF Learner

SoGood Talk: Lukas Pensel and Stefan Kramer. Forecast of Study Success in the STEM Disciplines Based Solely on Academic Record

The DeCoDeML workshop was organized for the first time and participation was well beyond the expectations of the organizing committee.

* = Johannes Gutenberg-Universität Mainz
^ = University of Birmingham