IIASS


Home 
spacer
Aims 
spacer
Lecturers 
spacer
Committees 
spacer
More Information 
spacer
Program 
spacer
Application 
spacer
Participants 
spacer
Site 
spacer
Ensemble Lab 

  

1 PROGRAM

Participants are expected to arrive in time for the evening meal on Sunday Sept 22th.

Monday  Sept 23th

9.00-9:30   Registration

Session 1 - Chair M. Marinaro, University  of Salerno, Italy and President of the IIAAS, Vietri, Italy.

9:30-10.40  Introduction to the Ensemble of Learning Machines -  N. Intrator,  Tel Aviv Univ., Israel and Brown Univ., Providence, RI, USA (references - slides)
10.40-11.00 Coffee break

11.00-12.20 Looking Inside of  the Black Box - L. Breiman, University of California at Berkeley, CA, USA (references - slides)

12.20-13.00 Introduction to the  Ensemble-Lab, organized by Francesco Masulli, University of Pisa, Italy (see here)

13.00-14.30 Lunch break

Session 2 - Chair  L. Breiman, University of California at Berkeley, CA, USA.

14.30-15.50 Ensemble Methods Based on Data Resampling - C. Furlanello, IRST, Trento, Italy  (references - slides)
15.50-16.10 Coffee break

16.10-17.30 ECOC Ensembles of  Learning Machines - F. Masulli, University of Pisa, Italy (references - slides)

17.30 -19.00  Ensemble-Lab  (see here)  /  Posters
 

Tuesday Sept 24th

Session 3 - Chair C. Furlanello, IRST, Trento, Italy.
9.00-10.20 Bias/Variance Analysis to Understand Ensemble Learning Algorithms -  T. Dietterich, Oregon State University, Oregon, USA  (references - slides)
10.20-10.40 Coffee break

10.40-12.00  Measures of Diversity in Classifier Combination (part 1) - L. Kuncheva, University of Wales, Bangor, UK (references  - slides)

12.00-14.00 Lunch break

Session 4 - Chair L. Reeker, NIST,  Gaithersburg, MD - USA. 

14.00-15.00 An Ensemble Method for Time Series Learning - F. Masulli, University of Pisa, Pisa, Italy (references  - slides)

15.00-15.50 Boosting Risk Classification of Geolocated Data - C. Furlanello, IRST, Trento, Italy   (references - slides)

15:50-16:10 Coffee break

16.10-17.30  Panel: Software  Tools  for Ensemble of Learning Machines organized by  N. Intrator,  Tel Aviv Univ., Israel and Brown Univ., Providence, RI, USA (references  - slides)

17.30-19.00   Ensemble-Lab  (see here)  / Posters

Wednesday Sept 25th

Session 5 - Chair J.M. DeLeo, National Institutes of Health Clinical Center, Bethesda, MD, USA. 
9.00-10.20 Geometrical Complexity of Classification Problems - T.K. Ho, Bell Laboratories, NJ, USA (references - slides)

10.20-10.40 Coffee break

10.40-12.00 Ensemble Methods in Classification of Remote Sensing  Images - L. Bruzzone, University of Trento, Trento, Italy (references  - slides)

12.00-14.00 Lunch break

Session 6 - Chair L. Kuncheva, University of Wales, Bangor, UK.

14.00-15.20  Linear Combiners for Fusion of Pattern Classifiers  - F. Roli, University of Cagliari, Cagliari, Italy (references - slides)
15:20-15:40 Coffee break

15.40-17.30  Participant talks (Posters 1, 2, 3, 4, 5, 6, 7, 8, 9)

17.30-19.00  Ensemble-Lab  (see here)  /  Posters

Thursday Sept 26th

Session 7 - Chair F. Roli, University of Cagliari, Cagliari, Italy.
9.00-10.20 Incremental Methods for MLP-RBF Networks -  N. Intrator,  Tel Aviv Univ., Israel and Brown Univ., Providence, RI, USA (references  - slides)

10.20-10.40 Coffee break

10.40-12.00 Algorithmic Stability and Model Selection for Bagging - M. Pontil, University of  Siena, Siena, Italy   (references  - slides)


Afternoon: Tour to  Paestum 
 
 
 

Friday Sept 27th

Session 8 - Chair  T. Dietterich, Oregon State University, Oregon, USA.  
9.00-10.20 Measures of Diversity in Classifier Combination (part 2) - L. Kuncheva, University of Wales, Bangor, UK (references  - slides)

10.20-10.40 Coffee break

10.40-12.00 Classifier Generating Methods and Stochastic Discrimination -  T.K. Ho, Bell Laboratories, NJ, USA (references  - slides)
12.00-14.00 Lunch break

Session 9 - Chair G. Gini, Politecnico di Milano, Milano, Italy.

14.00-15.20 Ensemble Learning Methods for Conditional Probability Estimation - T. Dietterich, Oregon State University, Oregon, USA  (references - slides)
15:20-15:40 Coffee break

15.40-17.00  Participant talks (Posters 10, 11, 12, 13, 14, 15, 16 )

17.00 -19.00  Ensemble-Lab  (see here)  /  Posters

Saturday Sept 28th

Session 10 - Chair T.K. Ho, Bell Laboratories, NJ, USA.
9.00-10.10 Ensemble Methods for Data Mining and Knowledge Extraction in Scientific Data Bases - G. Gini, Politecnico di Milano, Milano, Italy (references  - slides)

10.10-11.10 Ensemble Methods in Bioinformatics - G. Valentini, University of Genova, Italy (references  - slides)

11.10-11.30 Coffee break

Session 11 - Chairs N. Intrator,  Tel Aviv Univ., Israel and Brown Univ., Providence, RI, USA and F. Masulli, University of Pisa, Pisa, Italy.
11.30-12.40 Panel: Theoretical Perspectives and New Applications of  Ensemble of Learning Machines,  organized  by T. Dietterich, Oregon State University, Oregon, USA  (references - slides)
12.40-13.30 Participant talks on  results of the Ensemble-Lab (see here)

13.30-13.45 Conclusions 
 

END

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

2 POSTER / PARTICIPANT TALKS

 List of accept posters

Accepted posters will be displayed all the week 

Poster dimensions: length 90 cm - height 120 cm

Poster authors will  give a short oral presentation of  their works during the Participant talks of Wednesday Sept 25th  and Friday Sept 27th. Each presentation will be 10 min long. 

 

3 REFERENCES OF LECTURES


N. Intrator,  Introduction to the Ensemble of Learning Machines  (Monday  Sept 23) 

L. Breiman, Looking Inside of the Black Box (Monday  Sept 23) 

ftp://ftp.stat.berkeley.edu/pub/users/breiman/random-forests.abstract
ftp://ftp.stat.berkeley.edu/pub/users/breiman/
C. Furlanello,  Ensemble Methods Based on Data Resampling  (Monday  Sept 23) 
http://mpa.itc.it/furlan.html
http://www.research.att.com/~schapire/boost.html
F. Masulli, ECOC Ensembles of  Learning Machines (Monday  Sept 23)  
http://www.disi.unige.it/person/MasulliF/ricerca/publications.html#ens 


T. Dietterich, Bias/Variance Analysis to Understand Ensemble Learning Algorithms  (Tuesday Sept 24) 

  • S. Geman, E. Bienenstock and R. Doursat, "Neural Networks and the Bias/Variance Dilemma", Neural Computation 4, 1-58 (1992) [Not available on-line.]
  • Domingos, P. (2000). A Unified Bias-Variance Decomposition for Zero-One and Squared Loss. Proceedings of the Seventeenth National Conference on Artificial Intelligence (pp. 564-569), 2000. Austin, TX: AAAI http://www.cs.washington.edu/homes/pedrod/aaai00.ps.gz 
  • Valentini, G. and Dietterich, T. G. (2002). Bias-Variance Analysis and Ensembles of SVM. In J. Kittler and F. Roli (Ed.) Third International Workshop on Multiple Classifier Systems, Lecture Notes in Computer Science (pp. 222-231), 2002. New York: Springer Verlag.  ftp://ftp.cs.orst.edu/pub/tgd/papers/mcs2002-bv.ps.gz 
  • Dietterich, T. G. (2000). Ensemble Methods in Machine Learning, In J. Kittler and F. Roli (Ed.), First International Workshop on Multiple Classifier Systems, Lecture Notes in Computer Science (pp. 1-15). New York: Springer Verlag.  ftp://ftp.cs.orst.edu/pub/tgd/papers/mcs-ensembles.ps.gz 

    • Dietterich, T. G., (2000).  An experimental comparison of three methods for constructing ensembles of decision trees: Bagging, boosting, and randomization. Machine Learning, 40 (2) 139-158. ftp://ftp.cs.orst.edu/pub/tgd/papers/mlj-randomized-c4.ps.gz 


    L. Kuncheva, Measures of Diversity in Classifier Combination (part 1)   (Tuesday Sept 24)

    http://www.bangor.ac.uk/~mas00a/publications
    F.  Masulli,  An Ensemble Method for Time Series Learning (Tuesday Sept 24)
     http://www.disi.unige.it/person/MasulliF/ricerca/publications.html#tim 
    C. Furlanello, Boosting Risk Classification of Geolocated Data 
    (Tuesday Sept 24) 
    http://mpa.itc.it/furlan.html


    Panel: Software  Tools  for Ensemble of Learning Machines (Tuesday Sept 24) 

    T.K. Ho , Geometrical Complexity of Classification Problems  (Wednesday Sept 25)

  • Multiple Classifier Combination: Lessons and Next Steps, Tin Kam Ho, in A. Kandel, H. Bunke, (eds.), Hybrid Methods in Pattern Recognition, World Scientific, 2002.(.ps.gz) (.ps.zip) 
  • Complexity Measures of Supervised Classification Problems, Tin Kam Ho, Mitra Basu, IEEE Transactions on Pattern Analysis and Machine Intelligence,24 , 3, March 2002, 289-300. (.ps.gz) (.ps.zip) 
  • A Data Complexity Analysis of Comparative Advantages of Decision Forest Constructors, Tin Kam Ho, Pattern Analysis and Applications, to appear, 2002. (.ps.gz) (.ps.zip) 
  • The Learning Behavior of Single Neuron Classifiers on Linearly Separable or Nonseparable Input, Mitra Basu, Tin Kam Ho, Proceedings of the 1999 International Joint Conference on Neural Networks, Washington, D.C., July 10-16, 1999.(.ps.gz) (.ps.zip) 
    • L. Bruzzone, Ensemble Methods in Classification of Remote Sensing  Images (Wednesday Sept 25)
    http://www.ing.unitn.it/~bruzzone/RSLab/res_publ.html
    F. Roli, Fusion of Imbalanced Classifiers for Pattern Recognition Applications (Wednesday Sept 25)

    http://www.diee.unica.it/informatica/en/publications/publications-PattRecApps.html

    N. Intrator, Incremental Methods for MLP-RBF Networks  (Thursday Sept 26 )

    M. Pontil,  Algorithmic Stability and Model Selection for Bagging (Thursday Sept 26 )

    http://www.dii.unisi.it/~pontil/ 
    L. Kuncheva, Measures of Diversity in Classifier Combination (part 2) (Friday Sept 27) 
    http://www.bangor.ac.uk/~mas00a/publications
    L.K. Ho, Classifier Generating Methods and Stochastic Discrimination  (Friday Sept 27) 
    • Stochastic Discrimination, Eugune M. Kleinberg, Annals of Mathematics and Artificial Intelligence, 1, 1990, 207-239. 
    • An overtraining-resistant stochastic modeling method for pattern recognition, Eugene M. Kleinberg, Annals of Statistics, 4, 6, December 1996, 2319-2349. 
    • On the algorithmic implementation of stochastic discrimination, Eugene M. Kleinberg, IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-22, 5, May 2000, 473-490. 
    • A mathematically rigorous foundation for supervised learning, Eugeue M. Kleinberg, in J. Kittler, F. Roli, (eds.), Multiple Classifier Systems, Lecture Notes in Computer Science 1857, Springer, 2000, 67-76. 
    • Building Projectable Classifiers of Arbitrary Complexity, Tin Kam Ho, Eugene M. Kleinberg, Proceedings of the 13th International Conference on Pattern Recognition, Vienna, Austria, August 25-30, 1996, 880-885.(.ps.gz) (.ps.zip) 
    • The Random Subspace Method for Constructing Decision Forests, Tin Kam Ho, IEEE Transactions on Pattern Analysis and Machine Intelligence,20 , 8, August 1998, 832-844. (.ps.gz) (.ps.zip) 
    • Nearest Neighbors in Random Subspaces, Tin Kam Ho, Proceedings of the Second International Workshop on Statistical Techniques in Pattern Recognition, Sydney, Australia, August 11-13, 1998, 640-648.(.ps.gz) (.ps.zip)
    T. Dietterich, Ensemble Learning Methods for Conditional Probability Estimation  (Friday Sept 27) 
  • Margineantu, D., Dietterich, T. G. (2002) Improved Class Probability Estimates from Decision Tree Models. To appear in Lecture N\otes in Statistics ftp://ftp.cs.orst.edu/pub/tgd/papers/tr-msri-2002.pdf 
  • Provost, F. and P. Domingos. Tree Induction for Probability-based Rankings. To appear in Machine  Learning http://pages.stern.nyu.edu/~fprovost/Papers/pet-mlj-final.ps 


    • G. Gini, Ensemble Methods for Data Mining and Knowledge Extraction in Scientific Data Bases (Saturday Sept 28) 

     http://www.elet.polimi.it/upload/gini/vietri.htm
    G. Valentini, Ensemble Methods in Bioinformatics (Saturday Sept 28) 
    Panel: Theoretical Perspectives and New Applications of Ensemble of Learning Machines (Saturday Sept 28)