Publications
2009
1.
Alaiz-Rodríguez, Rocio; Japkowicz, Nathalie; Tischer, Peter
Visualizing High Dimensional Classifier Performance Data Book Section
En: Advances in Data Management, pp. 105–129, Springer Berlin Heidelberg Berlin, Heidelberg, 2009.
Resumen | Enlaces | BibTeX | Etiquetas: classifier evaluation, High-dimensional data, Visual Representation
@incollection{alaiz-rodriguez_visualizing_2009,
title = {Visualizing High Dimensional Classifier Performance Data},
author = {Rocio Alaiz-Rodríguez and Nathalie Japkowicz and Peter Tischer},
url = {https://link.springer.com/chapter/10.1007/978-3-642-02190-9_6},
year = {2009},
date = {2009-01-01},
booktitle = {Advances in Data Management},
pages = {105–129},
publisher = {Springer Berlin Heidelberg Berlin, Heidelberg},
abstract = {This paper addresses classifier performance evaluation, emphasizing the importance of analyzing high-dimensional data. By using visual representations of evaluation data, the approach leverages human visual capabilities to gain insights, interact with, and draw meaningful conclusions about classifiers and domains. The paper demonstrates how projection techniques from high-dimensional to lower-dimensional spaces enable exploratory analysis. This method is presented as a generalization of traditional evaluation methods based on point metrics, offering a more informative and interactive evaluation process. The framework allows users to study data from both a classifier and domain perspective, which is not possible with conventional methods.},
keywords = {classifier evaluation, High-dimensional data, Visual Representation},
pubstate = {published},
tppubtype = {incollection}
}
This paper addresses classifier performance evaluation, emphasizing the importance of analyzing high-dimensional data. By using visual representations of evaluation data, the approach leverages human visual capabilities to gain insights, interact with, and draw meaningful conclusions about classifiers and domains. The paper demonstrates how projection techniques from high-dimensional to lower-dimensional spaces enable exploratory analysis. This method is presented as a generalization of traditional evaluation methods based on point metrics, offering a more informative and interactive evaluation process. The framework allows users to study data from both a classifier and domain perspective, which is not possible with conventional methods.
2008
2.
Alaiz-Rodríguez, Rocio; Japkowicz, Nathalie; Tischer, Peter
Visualizing classifier performance on different domains Artículo de revista
En: 2008 20th IEEE International Conference on Tools with Artificial Intelligence, vol. 2, pp. 3–10, 2008, (Publisher: IEEE).
Resumen | Enlaces | BibTeX | Etiquetas: classifier evaluation, Visualization
@article{alaiz-rodriguez_visualizing_2008,
title = {Visualizing classifier performance on different domains},
author = {Rocio Alaiz-Rodríguez and Nathalie Japkowicz and Peter Tischer},
url = {https://ieeexplore.ieee.org/abstract/document/4669748},
year = {2008},
date = {2008-01-01},
journal = {2008 20th IEEE International Conference on Tools with Artificial Intelligence},
volume = {2},
pages = {3–10},
abstract = {Classifier performance evaluation typically gives rise to vast numbers of results that are difficult to interpret. On the one hand, a variety of different performance metrics can be applied; and on the other hand, evaluation must be conducted on multiple domains to get a clear view of the classifier's general behaviour. In this paper, we present a visualization technique that allows a user to study the results from a domain point of view and from a classifier point of view. We argue that classifier evaluation should be done on an exploratory basis. In particular, we suggest that, rather than pre-selecting a few metrics and domains to conduct our evaluation on, we should use as many metrics and domains as possible and mine the results of this study to draw valid and relevant knowledge about the behaviour of our algorithms. The technique presented in this paper will enable such a process.},
note = {Publisher: IEEE},
keywords = {classifier evaluation, Visualization},
pubstate = {published},
tppubtype = {article}
}
Classifier performance evaluation typically gives rise to vast numbers of results that are difficult to interpret. On the one hand, a variety of different performance metrics can be applied; and on the other hand, evaluation must be conducted on multiple domains to get a clear view of the classifier's general behaviour. In this paper, we present a visualization technique that allows a user to study the results from a domain point of view and from a classifier point of view. We argue that classifier evaluation should be done on an exploratory basis. In particular, we suggest that, rather than pre-selecting a few metrics and domains to conduct our evaluation on, we should use as many metrics and domains as possible and mine the results of this study to draw valid and relevant knowledge about the behaviour of our algorithms. The technique presented in this paper will enable such a process.
3.
Alaiz-Rodríguez, Rocío; Japkowicz, Nathalie; Tischer, Peter
A visualization-based exploratory technique for classifier comparison with respect to multiple metrics and multiple domains Artículo de revista
En: Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pp. 660–665, 2008, (Publisher: Springer Berlin Heidelberg Berlin, Heidelberg).
Resumen | Enlaces | BibTeX | Etiquetas: classifier evaluation, machine learning, model testing, performance metrics
@article{alaiz-rodriguez_visualization-based_2008,
title = {A visualization-based exploratory technique for classifier comparison with respect to multiple metrics and multiple domains},
author = {Rocío Alaiz-Rodríguez and Nathalie Japkowicz and Peter Tischer},
url = {https://link.springer.com/chapter/10.1007/978-3-540-87481-2_43},
year = {2008},
date = {2008-01-01},
journal = {Joint European Conference on Machine Learning and Knowledge Discovery in Databases},
pages = {660–665},
abstract = {Classifier performance evaluation typically gives rise to a multitude of results that are difficult to interpret. On the one hand, a variety of different performance metrics can be applied, each adding a little bit more information about the classifiers than the others; and on the other hand, evaluation must be conducted on multiple domains to get a clear view of the classifier’s general behaviour.},
note = {Publisher: Springer Berlin Heidelberg Berlin, Heidelberg},
keywords = {classifier evaluation, machine learning, model testing, performance metrics},
pubstate = {published},
tppubtype = {article}
}
Classifier performance evaluation typically gives rise to a multitude of results that are difficult to interpret. On the one hand, a variety of different performance metrics can be applied, each adding a little bit more information about the classifiers than the others; and on the other hand, evaluation must be conducted on multiple domains to get a clear view of the classifier’s general behaviour.