[Publications]


Spyridon Samothrakis
Research Fellow @ University of Essex
email: ssamot [at] essex [dot] ac [dot] uk
Academic Publications
[1] Spyridon Samothrakis, Tom Vodopivec, Michael Fairbank, and Maria Fasli. Convolutional-Match Networks for Question Answering. In International Joint Conference in Artificial Intelligence, August 2017. [ bib | .pdf ]
[2] Spyridon Samothrakis, Diego Perez, and Simon Lucas. Training gradient boosting machines using curve-fitting and information-theoretic features for causal direction detection. In Berna Bakır Batu, Isabelle Guyon, and Alexander Statnikov, editors, Springer Series on Challenges in Machine Learning : Cause-effect Pairs in Machine Learning. 2017. [ bib | .pdf ]
[3] Spyridon Samothrakis, Maria Fasli, Diego Perez, and Simon Lucas. Default policies for global optimisation of noisy functions with severe noise. Journal of Global Optimization, 67(4):893-907, 2017. [ bib | .pdf ]
[4] Spyridon Samothrakis. Narrative Progression Traits for Role-Playing Games. Game and Puzzle Design, July 2016. [ bib | .pdf ]
[5] Spyridon Samothrakis, Tom Vodopivec, Maria Fasli, and Michael Fairbank. Match Memory Recurrent Networks. In IEEE World Congress in Computational Intelligence, July 2016. [ bib | .pdf ]
[6] Diego Perez, Spyridon Samothrakis, Julian Togelius, Tom Schaul, Simon M. Lucas, Adrien Couetoux, Jerry Lee, Chong-U Lim, and Tommy Thompson. The 2014 General Video Game Playing Competition. IEEE Transactions on Computational Intelligence and AI in Games, 2015. [ bib ]
[7] Spyridon Samothrakis, Diego Perez-Liebana, Simon M Lucas, and Maria Fasli. Neuroevolution for general video game playing. In Computational Intelligence and Games (CIG), 2015 IEEE Conference on, pages 200-207. IEEE, 2015. [ bib ]
[8] Spyridon Samothrakis and Maria Fasli. Emotional sentence annotation helps predict fiction genre. PLoS ONE, 10(11):1-10, 11 2015. [ bib | DOI | http ]
[9] Diego Perez, Edward Powley, Daniel Whitehouse, Spyridon Samothrakis, Simon Lucas, and Peter I. Cowling. The 2013 multi-objective physical travelling salesman problem competition. In Proceedings of the IEEE Congress on Evolutionary Computation, 2014. [ bib ]
[10] Simon Lucas, Spyridon Samothrakis, and Diego Perez. Fast Evolutionary Adaptation for Monte Carlo Tree Search. In EvoGames, 2014. [ bib ]
[11] Diego Perez, Spyridon Samothrakis, and Simon Lucas. Knowledge-based Fast Evolutionary MCTS for General Video Game Playing. In Proceedings of the Conference on Computational Intelligence and Games (CIG), 2014. [ bib ]
[12] Diego Perez, Sanaz Mostaghim, Spyridon Samothrakis, and Simon Lucas. Multi-Objective Monte Carlo Tree Search for Real-Time Games. IEEE Transactions on Computational Intelligence and AI in Games, pp:1-13, 2014. [ bib ]
[13] Spyridon Samothrakis, Samuel A. Roberts, Diego Perez, and Simon Lucas. Rolling Horizon methods for Games with Continuous States and Actions. Proceedings of the Conference on Computational Intelligence and Games (CIG), Aug 2014. [ bib ]
[14] Spyridon Samothrakis, Diego Perez, Philipp Rohlfshagen, and Simon Lucas. Predicting Dominance Ranking for Score-based Games. Computational Intelligence and AI in Games, IEEE Transactions on, (Accepted), 2014. [ bib ]
[15] Diego Perez, Spyridon Samothrakis, and Simon Lucas. Online and offline learning in multi-objective monte carlo tree search. In Proceedings of the 8th International Conference on Computational Intelligence and Games, August 2013. [ bib ]
[16] Diego Perez, Spyridon Samothrakis, Simon Lucas, and Philipp Rohlfshagen. Rolling Horizon Evolution versus Tree Search for Navigation in Single-Player Real-Time Games. In Proceedings of the 15th Annual conference on Genetic and evolutionary computation,, GECCO '15, July 2013. [ bib ]
[17] Diego Perez, Spyridon Samothrakis, Simon Lucas, and Philipp Rohlfshagen. Rolling Horizon Evolution versus Tree Search for Navigation in Single-Player Real-Time Games. In Proceedings of the 15th Annual conference on Genetic and evolutionary computation,, GECCO '15, July 2013. [ bib ]
[18] Spyridon Samothrakis, Simon Lucas, Thomas Philip Runarsson, and David Robles. Coevolving Game-Playing Agents: Measuring Performance and Intransitivities. Evolutionary Computation, IEEE Transactions on, April 2013. There is a lot of of cycling in co-evolution, how do we get around it ?. [ bib | .pdf ]
[19] Diego Perez, Julian Togelius, Spyridon Samothrakis, Philipp Rohlfshagen, and Simon Lucas. Automated Map Generation for the Physical Travelling Salesman Problem. IEEE Transactions on Evolutionary Computation, pages 1-14, 2013. [ bib ]
[20] Diego Perez, Edward J. Powley, Daniel Whitehouse, Philipp Rohlfshagen, Spyridon Samothrakis, Peter I. Cowling, and Simon Lucas. Solving the Physical Travelling Salesman Problem: Tree Search and Macro-Actions. IEEE Transactions on Computational Intelligence and AI in Games, 6:1:31-45, 2013. [ bib ]
[21] Spyridon Samothrakis, Diego Perez, and Simon Lucas. Training gradient boosting machines using curve-fitting and information-theoretic features for causal direction detection. In NIPS 2013 Workshop on Causality, (3rd Prize in Kaggle Causality Challenge), 2013. [ bib ]
[22] Diego Perez, Edward J. Powley, Daniel Whitehouse, Philip Rohlfshagen, Spyridon Samothrakis, Peter I. Cowling, and Simon M. Lucas. Solving the physical travelling salesman problem: Tree search and macro-actions. Computational Intelligence and AI in Games, IEEE Transactions on,, 2013. [ bib ]
[23] Daniel Ashlock, Wendy Ashlock, Spyridon Samothrakis, Simon Lucas, and Colin Lee. From competition to cooperation: Co-evolution in a rewards continuum. In Proceedings of the 8th International Conference on Computational Intelligence and Games, September 2012. [ bib ]
[24] Karl Friston, Spyridon Samothrakis, and Read Montague. Active inference and agency: optimal control without cost functions. Biological Cybernetics, pages 1-19, August 2012. [ bib | .pdf ]
[25] C.B. Browne, E. Powley, D. Whitehouse, S.M. Lucas, P.I. Cowling, P. Rohlfshagen, S. Tavener, D. Perez, S. Samothrakis, and S. Colton. A survey of monte carlo tree search methods. Computational Intelligence and AI in Games, IEEE Transactions on, 4(1):1 -43, March 2012. A massive Survey of MCTS related papers. [ bib | .pdf ]
[26] Spyridon Samothrakis and Simon Lucas. Approximating n-player Behavioural Strategy Nash Equilibria Using Coevolution. In Proceedings of the 13th Annual conference on Genetic and evolutionary computation, GECCO '11, pages 293-300, July 2011. I have a much improved algorithm now, will be published some day somewhere. [ bib | .pdf ]
[27] Spyridon Samothrakis, David Robles, and Simon Lucas. Fast Approximate Max-n Monte-Carlo Tree Search for Ms Pac-Man. Computational Intelligence and AI in Games, IEEE Transactions on, 3(1):1 -16, June 2011. It could be that any model based method (when the opponent model is known) or maxn would perform better here, but at least MCTS provided a unified framework for attacking hunter-prey maze games. [ bib | .pdf ]
[28] Spyridon Samothrakis and Simon Lucas. Planning using online evolutionary overfitting. In Computational Intelligence (UKCI), 2010 UK Workshop on, September 2010. I am using a CTRNN here, instead of an MLP, although I am providing full and complete state as inputs. It's simply superfluous and makes the evolution and testing much slower. [ bib | .pdf ]
[29] Spyridon Samothrakis, David Robles, and Simon Lucas. A UCT agent for Tron: Initial investigations. In Proceedings of the 6th International Conference on Computational Intelligence and Games, August 2010. MCTS struggles with Tron (min-max performs better). I still don't know why, but I assume it's because MCTS does not work well with games of low branching factor. [ bib | .pdf ]