Dr. Günther A. Hoffmann Academic Research: Methods | Applications Other Interests Publications: Book Chapters | Papers Contact

Academic

• Humboldt University, Berlin Germany
  • Institut für Informatik / Institute of Computer Science
    Rechnerorganisation und Kommunikation (Prof. Malek)
• Duke University, North Carolina, USA
  • Pratt School of Engineering - Department of Electrical Engineering and Computer Science (Prof. Trivedi)
• Technical University, Berlin Germany
  • Institute of Systems Analysis (Prof. Krallmann)
  • Institute of Bionics and Evolution Strategy (Prof. Rechenberg)
• MIT - Massachusetts Institute of Technology, Massachusetts, USA
  • Department of Brain and Cognitive Sciences (Prof. Poggio)
  • Sloan School of Management (Prof. Lo)
• CCNY - City University of New York, New York, USA

Research Interests: Methods

• machine learning
  • quantitative, probabilistic, nonlinear, data driven modeling techniques
  • kernel techniques: radial basis functions (RBF), universal basis functions (UBF),
    support vector machines (SVM)
  • hidden markov models (HMM), multi layer perceptrons (MLP), projection pursuit (PP)
• stochastic optimization (Evolution Strategies, Genetic Algorithms) and Monte Carlo techniques
• function approximation
• feature detection, complexity reduction, variable selection
  • forward selection, backward elimination, probabilistic wrappers

Research Interests: Applications

• quantitative analysis of multivariate high frequency high volume data
• financial engineering, quantitative finance
  • derivative, bonds and stocks pricing
  • forex analysis
  • analyst evaluation
  • risk evaluation
  • ...
• fault tolerant computer systems: modeling and prediction of
  • dependability, availability, reliability
  • failures
  • resource requirements
  • ...
• self-* systems (e.g. self-managing, self-configuring, self-healing,...)
• natural language processing (NLP)
  • text mining from unstructured data
  • entity extraction, information extraction
  • semantic search
  • business intelligence, competitive intelligence

Other Interests

• skydiving, traveling
• technical entrepreneuship

Publications

Book Chapters

• Prediction-Based Software Availability Enhancement
  Salfner, Hoffmann, Malek (2005); in Lecture Notes in Computer Science, Volume 3460 / 2005, Title: Self-star Properties in Complex Information Systems: Conceptual and Practical Foundations, Editors: O. Babaoglu, M. Jelasity, A. Montresor, C. Fetzer, S. Leonardi, A. van Moorsel and M. van Steen, Springer-Verlag, Berlin 2005, ISBN: 3-540-26009-9
  
• Adaptive Transfer Functions in Radial Basis Function Networks
  Hoffmann (2004); in Lecture Notes in Computer Science LNCS by Springer; (ICCS2004), Volume 3037 / 2004, June 2004, Krakau, Eds.: Bubak M., Dick van Albada, Peter M.A. Sloot, Jack J. Dongarra, Springer-Verlag, Berlin 2004, ISBN: 3-540-22115-8
  
  

Papers

• A Best Practice Guide to Resources Forecasting for the Apache Webserver
  Hoffmann, Trivedi, Malek (2007); IEEE Transactions on Reliability, Dec. 2007 Volume: 56, Issue: 4 On page(s): 615-628, ISSN: 0018-9529
• A Best Practice Guide to Resource Forecasting for the Apache Webserver
  *BEST PAPER AWARD* Hoffmann, Trivedi, Malek (2006); 12th IEEE International Symposium Pacific Rim Dependable Computing (PRDC'06), University of California, Riverside, USA, December 18-20, 2006
• Call Availability Prediction in a Telecommunication System: A Data Driven Empirical Approach
  Hoffmann, Malek (2006); 25th IEEE Symposium on Reliable Distributed Systems (SRDS 2006), 2-4 October 2006 in Leeds, United Kingdom
• Invariant Detection for Apache Server Systems
  Hoffmann, Trivedi, Malek (2006); Duke University - Pratt School of Engineering, Humboldt-Universität zu Berlin Institut für Informatik
• Failure Prediction in Complex Computer Systems: A Probabilistic Approach
  Hoffmann (2005); Ph.D. Thesis, Humboldt-Universität zu Berlin Institut für Informatik
• Self–Rejuvenation - an Effective Way to High Availability
  Malek, Salfner, Hoffmann (2004); SELF-STAR:International Workshop on Self-Star Properties in Complex Information Systems , 31 May - 2 June 2004, University of Bologna Residential Center Bertinoro (Forlì), Italy
• Meeting Deadlines in Complex Systems
  Hoffmann, Malek (2004); Technical Report #171, Humboldt University
• Advanced failure prediction in a complex communication system
  Hoffmann, Salfner, Malek (2004); Technical Report #172, Humboldt University
• Data specific mixture kernels in Basis Function networks
  Hoffmann (2004) ; 11th International Conference on Neural Information Processing Systems (ICONIP) 2004, Calcutta
• A Probabilistic Approach to Wireless Network Modeling
  Hoffmann (1998); ICONIP ’98, 5th International Conference on Neural Information Processing, October 1998, Japan - Kitakyushu
• Evolving Parameter Sets for Conditional Density Models
  Hoffmann (1997) ; NOLTA - International Symposium on Nonlinear Theory and its Application; November 1997; Hawaii
• Experiments with Property Predictions of Distributed Computing Systems
  Hoffmann (1997) ; IASTED - Conference on Modeling and Simulation; May 1997; Pittsburgh
• A Radial Basis Function Approach to Modeling Resource Demands in Distributed Computing Systems
  Hoffmann (1996) ; ISMIP - International Symposium on Multi Technology Information Processing; September 1996; Taiwan
• Finding Good Information Sets: A Comparative Analysis
  Hoffmann (1996); NNCM - Neural Networks in the Capital Markets; November 1996; Pasadena
• Function Approximation with Learning Networks in the Financial Field and its Application to the Interest Rate Sector
  Hoffmann (1995) ; CIFER - Computational Intelligence for Financial Engineering; April 1995; New York
• Connectionist Forecasting: Modeling Financial Data with Neural Networks
  Hoffmann (1995); Master Thesis – Massachusetts Institute of Technology and Technical University of Berlin; January 1995; Boston - Berlin
  

Contact

Dr.-Ing. Günther A. Hoffmann

Humboldt-Universität zu Berlin
Institut für Informatik
Rechnerorganisation und Kommunikation
Unter den Linden 6
D-10099 Berlin
Germany
www.informatik.hu-berlin.de/rok/

Email: gunho ( at ) informatik.hu-berlin.de
Tel. +49 30 2093 3028
Fax +49 30 2093 3029
http://www.informatik.hu-berlin.de/~gunho/

Visitors
Johann von Neumann-Haus
Haus IV, 2. und 3. Etage
Rudower Chaussee 25
D-12489 Berlin
Lageplan http://www.wista.de/de/content/lage/3dkarte.htm

.