R
Robert P. W. Duin
Researcher at Delft University of Technology
Publications - 301
Citations - 32657
Robert P. W. Duin is an academic researcher from Delft University of Technology. The author has contributed to research in topics: Classifier (UML) & Random subspace method. The author has an hindex of 62, co-authored 301 publications receiving 31072 citations. Previous affiliations of Robert P. W. Duin include Utrecht University.
Papers
More filters
Journal ArticleDOI
The use of continuous variables for labeling objects
TL;DR: A summary is given of the various pattern recognition situations in which continuous variables may be used for labeling objects, and solutions are discussed and an example is given.
Book ChapterDOI
Towards Scalable Prototype Selection by Genetic Algorithms with Fast Criteria
Yenisel Plasencia-Calaña,Mauricio Orozco-Alzate,Heydi Méndez-Vázquez,Edel García-Reyes,Robert P. W. Duin +4 more
TL;DR: This paper proposes two methods based on genetic algorithms, one supervised and the other unsupervised, whose analyses provide an answer to the question: are genetic algorithms good for scalable prototype selection?
Proceedings Article
Possibilities of Zero-Error Recognition by Dissimilarity Representations
TL;DR: It is argued that this limitation does not hold for dissimilarity based representations for pattern recognition, and suggestions are made how this may be exploited in practice.
Proceedings Article
The Area under the ROC Curve as a Criterion for Clustering Evaluation
TL;DR: A measure that provides the robustness of an algorithm for several values of k, which constructs a ROC curve and measures the area under that curve, and it is shown that this measure can be used as a validation criterion in a semi-supervised context.
Journal ArticleDOI
Pattern recognition as a statistical method for analyzing the confidence level in the identification of objects
TL;DR: This paper gives a description of a statistical analysis procedure applied to concentration data obtained by non-destructive neutron activation analysis, when using this technique for the identification of flint artifacts.