Publicaties
Gekozen filters:
Gekozen filters:
An experimental approach for selection/elimination in stream network generalization using support vector machines Universiteit Gent
Automatic architectural style detection using one-class support vector machines and graph kernels Universiteit Gent
Bouligand derivatives and robustness of support vector machines for regression Vrije Universiteit Brussel
We investigate robustness properties for a broad class of support vector machines with non-smooth loss functions. These kernel methods are inspired by convex risk minimization in infinite dimensional Hilbert spaces. Leading examples are the support vector machine based on the eps-insensitive loss function, and kernel based quantile regression based on the pinball loss function. Firstly, we propose with the Bouligand influence function (BIF) a ...
Bouligand derivatives and robustness of support vector machines for regression Universiteit Gent
We investigate robustness properties for a broad class of support vector machines with non-smooth loss functions. These kernel methods are inspired by convex risk minimization in infinite dimensional Hilbert spaces. Leading examples are the support vector machine based on the e-insensitive loss function, and kernel based quantile regression based on the pinball loss function. Firstly, we propose with the Bouligand influence function (BIF) a ...
Multi-objective surrogate based optimization of gas cyclones using support vector machines and CFD simulations Vrije Universiteit Brussel
In order to accurately predict the complex non-linear relationships between the cyclone performance parameters (The Euler and Stokes numbers) and the four significant geometrical dimensions (the inlet section height and width, the vortex finder diameter and the cyclone total height), the support vector machines approach has been used. Two support vector regression surrogates (SVR) have been trained and tested by CFD data sets. The result ...
The potential of support vector machines and Kriging in modelling the gas cyclone performance Vrije Universiteit Brussel
The gas cyclone two performance parameters, the Euler and Stokes numbers, are highly affected by the variations in the geometrical parameters. To accurately predict the complex non-linear relationships between the cyclone performance and the seven geometrical parameters, the support vector machines approach has been applied. The support vector regression surrogate (SVR) has been trained and tested by an experimental dataset for the Euler number ...
A Review on Consistency and Robustness Properties of Support Vector Machines for Heavy-Tailed Distributions Vrije Universiteit Brussel
Support vector machines (SVMs) belong to the class of modern statistical machine learning techniques and can be described as M-estimatorswith aHilbert norm regularization term for functions. SVMs are consistent and robust for classification and regression purposes if based on a Lipschitz continuous loss and a bounded continuous kernel with a dense reproducing kernel Hilbert space. For regression, one of the conditions used is that the output ...
Named Entity Recognition in Biomedical Literature: A Comparison of Support Vector Machines and Conditional Random Fields Vrije Universiteit Brussel
In this paper, we propose two named entity recognition systems for biomedical literature, System1 using support vector machines and System2 using conditional random fields. Through employing several sets of experiments, we make a comprehensive comparison between these two systems. The final results reflect that System2 can achieve higher accuracy than System1, because System2 can catch more essential properties by handling the richer set of ...
A review on consistency and robustness properties of support vector machines for heavy-tailed distributions Universiteit Gent
Support vector machines (SVMs) belong to the class of modern statistical machine learning techniques and can be described as M-estimators with a Hilbert norm regularization term for functions. SVMs are consistent and robust for classification and regression purposes if based on a Lipschitz continuous loss and a bounded continuous kernel with a dense reproducing kernel Hilbert space. For regression, one of the conditions used is that the output ...