Using nonlinear dimensionality reduction to visualize classifiers

Authors:
Alexander Schulz;Andrej Gisbrecht;Barbara Hammer
Affiliations:
CITEC Centre of Excellence, University of Bielefeld, Germany;CITEC Centre of Excellence, University of Bielefeld, Germany;CITEC Centre of Excellence, University of Bielefeld, Germany
Venue:
IWANN'13 Proceedings of the 12th international conference on Artificial Neural Networks: advances in computational intelligence - Volume Part I
Year:
2013

Citing 14
Cited 0

Pairwise classification and support vector machines

Advances in kernel methods
Nomograms for visualizing support vector machines

Proceedings of the eleventh ACM SIGKDD international conference on Knowledge discovery in data mining
Improved learning of Riemannian metrics for exploratory analysis

Neural Networks - 2004 Special issue: New developments in self-organizing systems
Generalized Discriminant Analysis Using a Kernel Approach

Neural Computation
Unsupervised Learning of Image Manifolds by Semidefinite Programming

International Journal of Computer Vision
Kernel clustering-based discriminant analysis

Pattern Recognition
Visual Methods for Examining SVM Classifiers

Visual Data Mining
Nonlinear Dimensionality Reduction

Nonlinear Dimensionality Reduction
Information Retrieval Perspective to Nonlinear Dimensionality Reduction for Data Visualization

The Journal of Machine Learning Research
Interactive Data Visualization: Foundations, Techniques, and Applications

Interactive Data Visualization: Foundations, Techniques, and Applications
LIBSVM: A library for support vector machines

ACM Transactions on Intelligent Systems and Technology (TIST)
Fast track article: Looking ahead in pervasive computing: Challenges and opportunities in the era of cyber-physical convergence

Pervasive and Mobile Computing
SVMV – a novel algorithm for the visualization of SVM classification results

ISNN'06 Proceedings of the Third international conference on Advances in Neural Networks - Volume Part I
A general framework for dimensionality-reducing data visualization mapping

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Nonlinear dimensionality reduction (DR) techniques offer the possibility to visually inspect a given finite high-dimensional data set in two dimensions. In this contribution, we address the problem to visualize a trained classifier on top of these projections. We investigate the suitability of popular DR techniques for this purpose and we point out the benefit of integrating auxiliary information as provided by the classifier into the pipeline based on the Fisher information.