Dr. Patrick Traxler
F. Sobieczky, C. Lettner, T. Natschläger, P. Traxler. Adaptive heat pump and battery storage demand side energy management. E3S Web of Conferences, volume volume 22 (ASEE17), pages article number 00162, DOI 10.1051/e3sconf/20172200162, November, 2017.
A. Kogler, P. Traxler. Efficient and robust median-of-means algorithms for location and regression. Proceedings of the 18th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC 2016), pages online, DOI 10.1109/SYNASC.2016.041, January, 2017.
P. Traxler, A. Kogler. From fault detection to alarm lists - The photovoltaics use case. Poster at DX'17 - 28th International Workshop on Principles of Diagnosis, Brescia, Italy, September 26-29, 2017, September, 2017.
A. Kogler, P. Traxler. Health monitoring of large amounts of photovoltaic systems - A case study. In M. Indrawan-Santiago, I. Salvadori, M. Steinbauer, I. Khalil, G. Anderst-Kotsis (editors), Proceedings of the 19th International Conference on Information Integration and Web-based Applications & Services (iiWAS 2017), pages 385-389, DOI 10.1145/3151759.3151828, ACM, December, 2017.
A. Kogler, P. Traxler. Locating faults in photovoltaic systems data. In W. Woon, Z. Aung, O. Kramer, S. Madnick (editors), Data Analytics for Renewable Energy Integration - Proc. 4th ECML PKDD Workshop, DARE 2016, Lecture Notes in Artificial Intelligence, volume 10097, pages 1-9, DOI 10.1007/978-3-319-50947-1_1, January, 2017.
A. Kogler, P. Traxler. Parallel and robust empirical risk minimization via the median trick. In J. Blömer, I. Kotsiraes, T. Kutsia, D. Simos (editors), Mathematical Aspects of Computer and Information Sciences - Poc. MACIS 2017, Lecture Notes in Computer Science, volume 10693, pages 378-391, DOI 10.1007/978-3-319-72453-9_31, Springer, December, 2017.
A. Kinz, H. Biedermann, P. Traxler, B. Freudenthaler, J. Isopp, W. Schröder, A. Schlegl. Smart Maintenance - Ressourcenintelligente antizipative Instandhaltung durch Condition Monitoring, Datenanalyse und Störungsprognostik. volume 49, pages 20-24, January, 2016.
P. Traxler. The relative exponential time complexity of approximate counting satisfying assignments. Algorithmica (Special Issue IPEC 2014), volume 75, number 2, pages 339-362, DOI: 10.1007/s00453-016-0134-y, June, 2016.
P. Traxler, T. Grill, P. Gomez. A robust alternative to correlation networks for identifying fault systems. Proceedings of the 26th International Workshop on Principles of Diagnosis (DX-15), pages 11-18, August, 2015.
T. Steinmaurer, P. Traxler, M. Zwick, R. Stumptner, L. Christian. Combining stream processing engines and big data storages for data analysis. In T. Andreasen, H. Christiansen, J. Cubero, Z. Ras (editors), Foundations of Intelligent Systems - Proc. ISMIS 2014, Lecture Notes in Computer Science, volume 8502, pages 476-485, Springer, June, 2014.
P. Traxler, M. Zhariy. Efficient sensor placement in flow networks and sensor networks. DX'14 - 25th Edition of the International Workshop on Principles of Diagnosis, Graz, Austria, September 8-11, 2014, pages http://dx-2014.ist.tugraz.at/papers/DX14_Tue_PM_posters_paper7.pdf, September, 2014.
P. Traxler. The relative exponential time complexity of approximate counting satisfying assignments. IPEC 2014 - 9th International Symposium on Parameterized and Exact Computation, Breslau, Polen, September 10-12, 2014, September, 2014.
Dr. Patrick Traxler leads the research team for Data Analysis Algorithms at the SCCH. He studied computer science at TU Vienna and has received his doctoral degree from ETH Zurich in July 2010. From 2006 to 2009, he worked as a scientific assistant at the Institute of Theoretical Computer Science at ETH Zurich. His responsibilities comprised research in the area of algorithms and complexity and teaching assistance for algorithms and programming courses. From January 2011 to October 2012, he worked at the Institute of Ubiquitous Meteorology (UBIMET), an international meteorological company. He has been working at SCCH since November 2012.
The focus of his research lies in the areas of algorithms, diagnosis, and optimization. The goals of his research are the integration and application of these three disciplines. He approaches fault diagnosis and detection problems via machine learning and combines learning-based and knowledge-based methods. Combining learning-based and knowledge-based methods is a central concern of artificial intelligence. In addition, he works on learning algorithms with guaranteed robustness by designing and analyzing optimization algorithms. During his career, Dr. Traxler gathered experience with applying these approaches to application areas such as renewable energy and meteorology, analysis of technical systems and (discrete) manufacturing processes.
- A. Kogler, P. Traxler. Parallel Empirical Risk Minimization, submitted, 2017.
- A. Kogler, P. Traxler. Efficient and Robust Median-of-Medians Algorithms with Applications to Fault Detection, 18th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC2016), 2016.
- P. Traxler, T. Grill, P. Gomez. A Robust Alternative to Correlation Networks for Identifying Faulty Systems, 26th International Workshop on Principles of Diagnosis, 2015.
- P. Traxler. The Relative Exponential Time Complexity of Approximate Counting Satisfying Assignments, to appear in Algorithmica (Special Issue), 2016. An extended abstract appeared in 9th International Symposium on Parameterized and Exact Computation, 2014.
- P. Traxler. Variable Influences in Conjunctive Normal Forms, 12th International Conference Theory and Applications of Satisfiability Testing, pages 101-113, 2009.
Selected research projects
- Since 2015, Jan.: COMET-project inFADIA
- Since 2014, Sept.: EStore-M (EraSME; FFG Nr. 836684)
- Since 2013, Sept.: Smart Maintenance (FFG Nr. 843650)
- 2012, Jan. - 2014, Dec.: COMET-project modDiscovery