Machine Learning and Data Mining in Pattern Recognition

1. Front Matter

   PDF

2. Invited Talk

   1. Data Clustering: User’s Dilemma

      • Anil K. Jain

      Pages 1-1

3. Classification

   1. On Concentration of Discrete Distributions with Applications to Supervised Learning of Classifiers

      • Magnus Ekdahl, Timo Koski

      Pages 2-16

   2. Comparison of a Novel Combined ECOC Strategy with Different Multiclass Algorithms Together with Parameter Optimization Methods

      • Marco Hülsmann, Christoph M. Friedrich

      Pages 17-31

   3. Multi-source Data Modelling: Integrating Related Data to Improve Model Performance

      • Paul R. Trundle, Daniel C. Neagu, Qasim Chaudhry

      Pages 32-46

   4. An Empirical Comparison of Ideal and Empirical ROC-Based Reject Rules

      • Claudio Marrocco, Mario Molinara, Francesco Tortorella

      Pages 47-60

   5. Outlier Detection with Kernel Density Functions

      • Longin Jan Latecki, Aleksandar Lazarevic, Dragoljub Pokrajac

      Pages 61-75

   6. Generic Probability Density Function Reconstruction for Randomization in Privacy-Preserving Data Mining

      • Vincent Yan Fu Tan, See-Kiong Ng

      Pages 76-90

   7. An Incremental Fuzzy Decision Tree Classification Method for Mining Data Streams

      • Tao Wang, Zhoujun Li, Yuejin Yan, Huowang Chen

      Pages 91-103

   8. On the Combination of Locally Optimal Pairwise Classifiers

      • Gero Szepannek, Bernd Bischl, Claus Weihs

      Pages 104-116

4. Feature Selection, Extraction and Dimensionality Reduction

   1. An Agent-Based Approach to the Multiple-Objective Selection of Reference Vectors

      • Ireneusz Czarnowski, Piotr Jȩdrzejowicz

      Pages 117-130

   2. On Applying Dimension Reduction for Multi-labeled Problems

      • Moonhwi Lee, Cheong Hee Park

      Pages 131-143

   3. Nonlinear Feature Selection by Relevance Feature Vector Machine

      • Haibin Cheng, Haifeng Chen, Guofei Jiang, Kenji Yoshihira

      Pages 144-159

   4. Affine Feature Extraction: A Generalization of the Fukunaga-Koontz Transformation

      • Wenbo Cao, Robert Haralick

      Pages 160-173

5. Clustering

   1. A Bounded Index for Cluster Validity

      • Sandro Saitta, Benny Raphael, Ian F. C. Smith

      Pages 174-187

   2. Varying Density Spatial Clustering Based on a Hierarchical Tree

      • Xuegang Hu, Dongbo Wang, Xindong Wu

      Pages 188-202

   3. Kernel MDL to Determine the Number of Clusters

      • Ivan O Kyrgyzov, Olexiy O Kyrgyzov, Henri Maître, Marine Campedel

      Pages 203-217

   4. Critical Scale for Unsupervised Cluster Discovery

      • Tomoya Sakai, Atsushi Imiya, Takuto Komazaki, Shiomu Hama

      Pages 218-232

   5. Minimum Information Loss Cluster Analysis for Categorical Data

      • Jiří Grim, Jan Hora

      Pages 233-247

   6. A Clustering Algorithm Based on Generalized Stars

      • Airel Pérez Suárez, José E. Medina Pagola

      Pages 248-262

MLDM / ICDM Medaillie Meissner Porcellan, the “White Gold” of King August the Strongest of Saxonia Gottfried Wilhelm von Leibniz, the great mathematician and son of Leipzig, was watching over us during our event in Machine Learning and Data Mining in Pattern Recognition (MLDM 2007). He can be proud of what we have achieved in this area so far. We had a great research program this year. This was the fifth MLDM in Pattern Recognition event held in Leipzig (www.mldm.de). Today, there are many international meetings carrying the title machine learning and data mining, whose topics are text mining, knowledge discovery, and applications. This meeting from the very first event has focused on aspects of machine learning and data mining in pattern recognition problems. We planned to reorganize classical and well-established pattern recognition paradigms from the view points of machine learning and data mining. Although it was a challenging program in the late 1990s, the idea has provided new starting points in pattern recognition and has influenced other areas such as cognitive computer vision. For this edition, the Program Committee received 258 submissions from 37 countries (see Fig. 1). To handle this high number of papers was a big challenge for the reviewers. Every paper was thoroughly reviewed and all authors received a detailed report on their submitted work.

• Spam
• classification
• cognition
• data mining
• learning
• machine learning
• pattern recognition

Policies and ethics