Remco Bouckaert

remco@cs.waikato.ac.nz

Recent publications

• Remco R. Bouckaert. DensiTree: Making Sense of Sets of Phylogenetic Trees. Bioinformatics 2010; doi: 10.1093/bioinformatics/btq110 (accompanying software and manual)
• Remco R. Bouckaert: A Hierarchical Face Recognition Algorithm. ACML 2009: 38-50
• Eibe Frank, Remco R. Bouckaert: Conditional Density Estimation with Class Probability Estimators. ACML 2009: 65-81
• R. Bouckaert, G. Holmes, B. Pfahringer and D. Fletcher. Gaussian Processes on Graphics Cards for NIR. NIR 2009
• B. Pfahringer, D. Fletcher, R.Bouckaert, G.Holmes Random Model Trees: a competitive off-the-shelf technology for NIR NIR 2009
• Remco R. Bouckaert: Practical Bias Variance Decomposition In Proc 21st Australasian Joint Conference on Artificial Intelligence Auckland, New Zealand, 2008 pdf
• Remco R. Bouckaert, Milan Studeny: Racing algorithms for conditional independence inference. Int. J. Approx. Reasoning 45(2): 386-401 (2007) pdf
• Remco R. Bouckaert, Michael Goebel, Patricia J. Riddle: Generalized Unified Decomposition of Ensemble Loss. Australian Conference on Artificial Intelligence 2006: 1133-1139 pdf extended version
• Remco R. Bouckaert: Efficient AUC Learning Curve Calculation. Australian Conference on Artificial Intelligence 2006: 181-191 pdf
• Remco R. Bouckaert: Voting Massive Collections of Bayesian Network Classifiers for Data Streams. Australian Conference on Artificial Intelligence 2006: 243-252 pdf
• Remco R. Bouckaert. Low Replicability of Machine Learning Experiments is not a Small Data Set Phenomenon. ICML Meta Learning workshop, 2005. pdf
• Remco R. Bouckaert and Milan Studeny. Racing for conditional independence inference. ECSQARU 221-232, 2005. Demonstration applets. pdf.
• Remco R. Bouckaert, Bayesian networks in Weka. Technical Report 14/2004. Computer Science Department. University of Waikato. 2004. pdf
• Remco R. Bouckaert, Naive Bayes Classifiers that Perform Well with Continuous Variables. In Proceedings of the 17th Australian Conference on AI (AI 04), Lecture Notes in AI, Springer. 2004. Extended report version
• Remco R. Bouckaert, Estimating Replicability of Classifier Learning Experiments. ICML, 2004. ps pdf
• Remco R. Bouckaert and Eibe Frank, Evaluating the Replicability of Significance Tests for Comparing Learning Algorithms. PAKDD, 2004. ps
• Remco R. Bouckaert, Choosing between two learning algorithms based on calibrated tests, International Conference on Machine Learning, 51--58, 2003.
  Conference version
  Extended report version
  Data used for the experiments
• Remco R. Bouckaert, A Probabilistic Line Breaking Algorithm, The 16th Australian Joint Conference on Artificial Intelligence - AI.03, 2003.
  pdf Extended report version
• Remco R. Bouckaert, Choosing learning algorithms using sign tests with high replicability, The 16th Australian Joint Conference on Artificial Intelligence - AI.03, 2003.postscript
• Accuracy bounds for ensembles under 0 - 1 loss, Remco R. Bouckaert, Working Paper 02/04, Computer science department, University of Waikato postscript
• R. R. Bouckaert. Low level information extraction: a Bayesian network based approach. TextML 2002. pdf
• Tomas Kocka, Remco R. Bouckaert, Milan Studeny. On characterizing Inclusion of Bayesian Networks. Uncertainty in artificial intelligence. 2001. pdf
• M. Studeny, R.R. Bouckaert, T. Kocka: Extreme supermodular set functions over five variables. Research report n. 1977, Institute of Information Theory and Automation, Prague, January 2000 (32 pages). postscript
• M. Studeny and R. R. Bouckaert: On chain graph models for description of conditional independence structures. The Annals of Statistics 26 (1998), n. 4, pp. 1434-1495. postscript

Working on...

Erdos number: 3

3: Enrique Castillo, Remco R. Bouckaert, Jose M. Sarabia, Cristina Solares: Error Estimation in Approximate Bayesian Belief Network Inference. UAI 1995: 55-62

2: E. Castillo and J. Galambos. Lifetime regression models based on a functional equation of physical nature. Journal of Applied Probability. 24:160-169, 1987.

1: Erdos, Paul; Galambos, Janos , Asymptotic distribution of normalized arithmetical functions. Proc. Am. Math. Soc. 46, 1-8 (1974).

GPU programming

Matrix inverse with CUDA and Cublas. Invert a 10.000x10.000 symmetric positive definite matrix in under 15 seconds on a NVidia GTX 280 card (14.38 seconds will do). source code . Also contain matrix inversion through Gaussian elimination for nxn matrices.

Other open source projects

BEAST 2: MCMC and evolution libraries for Bayesian phylogenetic analysis.

Snap: a Beast 2 add on for inferring species trees directly from SNP and AFLP data.

Xtal Mountain

Check out ShoXS, a Shorter XSL Syntax.

trip2009