Alexander Dockhorn
Dr. Alexander Dockhorn
Chair of Computational Intelligence
I studied Computer Science at the Otto-von-Guericke University (OvGU) in Magdeburg (Germany) and University of Abertay in Dundee (Scotland). Back in Magdeburg I finished my Master's degree in October 2015. My current research interests are computational intelligence in games and intelligent data analysis. My focus therein lies on partial information games, and learning rules and strategies from succesful replays. Additionally, I am always looking for real-world applications for computational intelligence methods.
My current research projects include, but are not limited to:
Forward Model Approximation
Part of my PhD project focuses on developing a process of modelling environment dynamics by observation. This would enable the application of statistical forward planning methods, which proved to be successful in a wide range of game playing scenarios. While my previous studies focused on the applications to general game learning similar methods could be used in the context of robotics and optimisation.
State Induction
In the second part of my PhD project I analyse how statistical forward planning methods can be applied to partial information games. New techniques for inducing the current state are analysed and compared to the state-of-the-art. Applications such as collectable card games present a natural benchmark and lead me to the creation of an international research competition on Hearthstone AI.
Hearthstone AI Competition
I am the organiser of the international research competition on Hearthstone AI, which was part of the IEEE Conference on Computational Intelligence and Games 2018. The competition focuses on the development of autonomous Hearthstone agents and tested the agent's skill on multiple game playing tasks. It attracted a total of 50 submissions and was the largest competition of the conference. Due to numerous requests, it will be held again at the IEEE Conference on Games 2019. More information can be found at http://www.ci.ovgu.de/Research/HearthstoneAI.html
Journal Papers
- Pascal Held, Alexander Dockhorn, and Rudolf Kruse
- On Merging and Dividing Social Graphs
- Journal of Artificial Intelligence and Soft Computing Research, 5(1) --> Link (Download)
Conference Papers
- Alexander Dockhorn, Simon M. Lucas, Vanessa Volz, Ivan Bravi, Raluca D. Gaina, and Diego Perez-Liebana
- Learning Local Forward Models on Unforgiving Games
-
Accepted for publication at the IEEE Conference on Games (CoG) 2019
- Simon M. Lucas, Alexander Dockhorn, Vanessa Volz, Chris Bamford, Raluca D. Gaina, Ivan Bravi, Diego Perez-Liebana, Sanaz Mostaghim, and Rudolf Kruse
- A Local Approach to Forward Model Learning: Results on the Game of Life Game
-
Accepted for publication at the IEEE Conference on Games (CoG) 2019
- Alexander Dockhorn, Tony Schwensfeier, and Rudolf Kruse
- Fuzzy Multiset Clustering for Metagame Analysis
-
Accepted for publication at the 11th Conference of the European Society for Fuzzy Logic and Technology, Prague, Czech Republic --> PDF, Github
- Alexander Dockhorn and Rudolf Kruse
- Detecting Sensor Dependencies for Building Complementary Model Ensembles
-
In Proceedings. 28. Workshop Computational Intelligence, Dortmund 2018 --> Link
- Alexander Dockhorn, Tim Tippelt, and Rudolf Kruse
- Model Decomposition for Forward Model Approximation
-
In 2018 IEEE Symposium Series on Computational Intelligence (SSCI)--> Link
- Alexander Dockhorn and Daan Apeldoorn
- Forward Model Approximation for General Video Game Learning
- In IEEE Computational Intelligence in Games (CIG) 2018 in Maastricht--> Link
- Alexander Dockhorn, Max Frick, Ünal Akkaya, and Rudolf Kruse
- Predicting Opponent's Moves for Improving Hearthstone AI
- 17th International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems (IPMU 2018)--> Link
- Alexander Dockhorn, Christoph Doell, Matthias Hewelt, and Rudolf Kruse
- A Decision Heuristic for MCTS Doppelkopf Agents
- In 2017 IEEE Symposium Series on Computational Intelligence (SSCI) in Hawaii--> Link
- Tim Sabsch, Christian Braune, Alexander Dockhorn, and Rudolf Kruse
- Using a Multiobjective Genetic Algorithm for Curve Approximation
- In 2017 IEEE Symposium Series on Computational Intelligence (SSCI) in Hawaii--> Link
- Alexander Dockhorn and Rudolf Kruse
- Combining Cooperative and Adversarial Coevolution in the Context of Pac-Man
- In IEEE Computational Intelligence in Games 2017 in New York --> Link
- Alexander Dockhorn, Christian Braune, and Rudolf Kruse
- Variable density based clustering
- In 2016 IEEE Symposium Series on Computational Intelligence (SSCI) in Athen--> Link
- Alexander Dockhorn, Christian Braune, and Rudolf Kruse
- An Alternating Optimization Approach based on Hierarchical Adaptations of DBSCAN
- In 2015 IEEE Symposium Series on Computational Intelligence (SSCI) in Capetown --> Link
- Pascal Held, Alexander Dockhorn, Rudolf Kruse
- Clustering Social Networks Using Competing Ant Hives
- In 2015 Second European Network Intelligence Conference --> Link
- Pascal Held, Alexander Dockhorn, Rudolf Kruse
- On Merging and Dividing of Barabasi-Albert-graphs
- In 2014 IEEE Symposium on Evolving and Autonomous Learning Systems (EALS) --> Link
- Pascal Held, Alexander Dockhorn, Rudolf Kruse
- Generating Events for Dynamic Social Network Simulations
- 15th International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems --> Link
Open Access Publications
- Alexander Dockhorn and Sanaz Mostaghim
- Introducing the Hearthstone-AI Competition
-
Open Access Paper on arXiv.org
Theses
- Alexander Dockhorn
- Density Based Clustering Optimization
- Master Thesis, Otto-von-Guericke University Magdeburg (OvGU), 2015
- Alexander Dockhorn
- Computergestützte Analyse onkologischer Daten mithilfe GraphischerModelle
- Bachelor Thesis, Otto-von-Guericke University Magdeburg (OvGU), 2014
Lectures and Seminars
- Tutorials for the lecture Fuzzy Systems, Summer 2019
- Tutorials for the lecture Computational Intelligence in Games, Summer 2019
- Tutorials for the lecture Bayes Networks, Winter 2018/2019
- Lecture Computational Intelligence in Games, Summer 2018
- Tutorials for the lecture Neural Networks, Summer 2018
- Seminar Classification Algorithms, Winter 2017/18
- Tutorials for the lecture Bayes Networks, Winter 2017/2018
- Tutorials for the lecture Computational Intelligence in Games, Summer 2017
- Seminar Deep Learning, Summer 2016
- Tutorials for the lecture Intelligente Systeme, Winter 2015/2016
- Tutorials for the lecture Bayes Networks, Winter 2015/2016
- Tutorials for the lecture Neural Networks, Summer 2015
- Tutorials for the lecture Intelligente Systeme, Winter 2014/2015
- Tutorial for the lecture Einführung in die Informatik, Winter 2011/2012
Advisor of Bachelor and Master Theses
Topics under work:
- T. T., Verwendung von Multi-Agenten Systemen zur Verhaltensmodellierung in unbekannten Spielen, Master Thesis
- T. S., Clustering Analysis of the Hearthstone Meta-Game, Master Thesis
- M. K., Kinematic Path Planning using Neural Network based Rollout Strategies, Master Thesis,
- S. D., Multivariate Time Series Sensor Data Clustering, Master Thesis
- B. H., Developing agents for pre-made Hearthstone decks, Bachelor Thesis
Finished:
- Pankaj Narula
- Designing an Interface between Data System and Root Cause Evaluation to enhance analysis strategy utilizing a larger Database for Tire Design, Manufacturing and Evaluation Process
- Master Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- December, 2018
- Jannis Becke
- Optimising All-Shortest-Path Dictionaries using Machine Learning
- Master Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- July, 2018
- Chris Saxton
- Ermitteln der Siegbedingung und Regeln unbekannter Spiele durch Assoziationsanalyse
- Bachelor Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- June, 2018
- Sujan Adhikari
- Regression Analysis for Power Consumption of a Production Plant
- Bachelor Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- May, 2018
- Cornelius Styp von Rekowski
- Robust and Transferable Reflectance Reconstruction Using Deep Neural Networks
- Bachelor Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- April, 2018
- Maurice Hoffmeister
- Multikriterielle Wegfindung für Agentengruppen
- Bachelor Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- April, 2018
- Matthias Hewelt
- Entscheidungsheuristische Erweiterung des UCT-Algorithmus für Doppelkopf
- Master Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- March, 2017
- Leopold Ryll
- Kursverlaufvorhersage vonWertpapierenmit Hilfe von Ensemble Classification
- Bachelor Thesis, OVGU
- Supervision: Rudolf Kruse
- Advisor: Alexander Dockhorn
- March, 2017