Papers by Steven De Bruyne
Lecture Notes in Computer Science, 2008
We investigate the effects of dimensionality reduction using different techniques and different d... more We investigate the effects of dimensionality reduction using different techniques and different dimensions on six two-class data sets with numerical attributes as pre-processing for two classification algorithms. Besides reducing the dimensionality with the use of principal components and linear discriminants, we also introduce four new techniques. After this dimensionality reduction two algorithms are applied. The first algorithm takes advantage of the reduced dimensionality itself while the second one directly exploits the dimensional ...
This paper reports on the construction of a personalized theme creation engine as a possible cata... more This paper reports on the construction of a personalized theme creation engine as a possible catalyst to the active use in secondary educationin Europe of digital media published on-line by selected museums
Lecture Notes in Computer Science, 2010
ABSTRACT Data visualisation can be a great support to the data mining process. We introduce a dat... more ABSTRACT Data visualisation can be a great support to the data mining process. We introduce a data structure that allows browsing through the data giving a complete but very manageable overview over the entire data set, where the data is split into subsets and displayed from interesting angles to reveal the relevant patterns for each subset. Based on the features originating from principal separation analysis, a tree is grown. A node of the tree is associated with a feature and a subset of instances, and later on with a two-dimensional visualisation. At the node level, groups of instances of different classes that can be displayed from a more interesting angle are temporarily grouped together in subsets. For each of these subsets child nodes are created that display this part of the data from a more interesting angle, revealing new patterns. This process is continued until no further improved visualisation can be found. After the tree has been constructed, it can be used to easily browse through the data. The nodes correspond with two-dimensional visualisations of the data, but the specific properties of the tree allow for three-dimensional animated transitions from one node to another, further clarifying the patterns in the data.
Annals of Operations Research, 2010
We consider the linear classification method consisting of separating two sets of points in d-spa... more We consider the linear classification method consisting of separating two sets of points in d-space by a hyperplane. We wish to determine the hyperplane which minimises the sum of distances from all misclassified points to the hyperplane. To this end two local descent methods are developed, one grid-based and one optimisation-theory based, and are embedded in several ways into a VNS metaheuristic scheme. Computational results show these approaches to be complementary, leading to a single hybrid VNS strategy which combines both approaches to exploit the strong points of each. Extensive computational tests show that the resulting method performs well.
We consider the linear classification method consisting of separating two sets of points in d-spa... more We consider the linear classification method consisting of separating two sets of points in d-space by a hyperplane. We investigate the situation where the two sets are nonseparable, and we wish to find the hyperplane which minimises the sum of distances from all misclassified points to the hyperplane. To this end two local descent methods are developed, one grid-based and one optimisation-theory based, and are embedded in several ways into a VNS metaheuristic scheme. Computational results show these approaches to be complementary, leading to a single hybrid VNS strategy which combines both approaches to exploit the strong points of each. Extensive computational tests show that the resulting method performs well.
We investigate the eects of dimensionality reduction using dierent techniques and dierent dimensi... more We investigate the eects of dimensionality reduction using dierent techniques and dierent dimensions on six two-class data sets with numerical attributes as pre-processing for two classification algo- rithms. Besides reducing the dimensionality with the use of principal components and linear discriminants, we also introduce four new tech- niques. After this dimensionality reduction two algorithms are applied. The first algorithm takes advantage of the reduced dimensionality itself while the second one directly exploits the dimensional ranking. We ob- serve that neither a single superior dimensionality reduction technique nor a straightforward way to select the optimal dimension can be iden- tified. On the other hand we show that a good choice of technique and dimension can have a major impact on the classification power, gen- erating classifiers that can rival industry standards. We conclude that dimensionality reduction should not only be used for visualisation or as pre-processing...
Uploads
Papers by Steven De Bruyne