Sebastian Widz - Academia.edu (original) (raw)
Papers by Sebastian Widz
A Rough Set-Based Magnetic Resonance Imaging Partial Volume Detection System
Lecture Notes in Computer Science, 2005
Application of rough set based dynamic parameter optimization to MRI segmentation
IEEE Annual Meeting of the Fuzzy Information, 2004. Processing NAFIPS '04., 2004
Springer eBooks, 2010
We propose a framework for experimental verification whether mechanisms of voting among rough-set... more We propose a framework for experimental verification whether mechanisms of voting among rough-set-based classifiers and criteria for extracting those classifiers from data should follow analogous mathematical principles. Moreover, we show that some of types of criteria perform better for high-quality data while the others are useful rather for low-quality data. The framework is based on the principles of approximate attribute reduction and probabilistic extensions of rough-set-based approach to data analysis. The framework is not supposed to produce the best-ever classification results, unless it is extended by some additional parameters known from the literature. Instead, our major goal is to illustrate in a possibly simplistic way that it is worth unifying mathematical background for the stages of learning and applying rough-set-based classifiers.
Evolutionary inspired optimization of feature subset ensembles
Rough-Set-Inspired Feature Subset Selection, Classifier Construction, and Rule Aggregation
Lecture Notes in Computer Science, 2011
ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from da... more ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from data. Each of feature subsets yields a single classifier, basically by generating its corresponding if-then decision rules from the training data. Feature subsets are extracted according to a simple randomized algorithm, following the filter (rather than wrapper or embedded) methodology. Classifier ensemble is built from single classifiers by defining aggregation laws on top of decision rules. We investigate whether rough-set-inspired methods can help in the steps of formulating feature subset optimization criteria, feature subset search heuristics, and the strategies of voting among classifiers. Comparing to our previous research, we pay a special attention to synchronization of the filter-based criteria for feature subset selection and extraction of rules basing on the obtained feature subsets. The overall framework is not supposed to produce the best-ever classification results, unless it is extended by some additional techniques known from the literature. Our major goal is to illustrate in a possibly simplistic way some general interactions between the above-mentioned criteria.
Approximation Degrees in Decision Reduct-Based MRI Segmentation
Rough-Set-based classifier extraction and voting criteria are Applied together?
Lecture Notes in Computer Science, 2010
Set Based Dynamic Parameter to MRI Segmentation
Lecture Notes in Computer Science, 2010
We propose a framework for experimental verification whether mechanisms of voting among rough-set... more We propose a framework for experimental verification whether mechanisms of voting among rough-set-based classifiers and criteria for extracting those classifiers from data should follow analogous mathematical principles. Moreover, we show that some of types of criteria perform better for high-quality data while the others are useful rather for low-quality data. The framework is based on the principles of approximate attribute reduction and probabilistic extensions of rough-set-based approach to data analysis. The framework is not supposed to produce the best-ever classification results, unless it is extended by some additional parameters known from the literature. Instead, our major goal is to illustrate in a possibly simplistic way that it is worth unifying mathematical background for the stages of learning and applying rough-set-based classifiers.
A Hybrid Approach to MR Imaging Segmentation Using Unsupervised Clustering and Approximate Reducts
Lecture Notes in Computer Science, 2005
An Automated Multi-spectral MRI Segmentation Algorithm Using Approximate Reducts
Lecture Notes in Computer Science, 2004
A Rough Set-Based Magnetic Resonance Imaging Partial Volume Detection System
Lecture Notes in Computer Science, 2005
Application of rough set based dynamic parameter optimization to MRI segmentation
IEEE Annual Meeting of the Fuzzy Information, 2004. Processing NAFIPS '04., 2004
Rough-Set-Inspired Feature Subset Selection, Classifier Construction, and Rule Aggregation
Lecture Notes in Computer Science, 2011
ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from da... more ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from data. Each of feature subsets yields a single classifier, basically by generating its corresponding if-then decision rules from the training data. Feature subsets are extracted according to a simple randomized algorithm, following the filter (rather than wrapper or embedded) methodology. Classifier ensemble is built from single classifiers by defining aggregation laws on top of decision rules. We investigate whether rough-set-inspired methods can help in the steps of formulating feature subset optimization criteria, feature subset search heuristics, and the strategies of voting among classifiers. Comparing to our previous research, we pay a special attention to synchronization of the filter-based criteria for feature subset selection and extraction of rules basing on the obtained feature subsets. The overall framework is not supposed to produce the best-ever classification results, unless it is extended by some additional techniques known from the literature. Our major goal is to illustrate in a possibly simplistic way some general interactions between the above-mentioned criteria.
Approximation Degrees in Decision Reduct-Based MRI Segmentation
2007 Frontiers in the Convergence of Bioscience and Information Technologies, 2007
Evolutionary inspired optimization of feature subset ensembles
2010 Second World Congress on Nature and Biologically Inspired Computing (NaBIC), 2010
ABSTRACT
Rough Set Based Decision Support—Models Easy to Interpret
Advanced Information and Knowledge Processing, 2012
Rapid evolution of technology allows people to record more data than ever. Gathered information i... more Rapid evolution of technology allows people to record more data than ever. Gathered information is intensively used by data analysts and domain experts. Collections of patterns extracted from data compose models (compact representations of discovered knowledge), which are at the heart of each decision support system. Models based on mathematically sophisticated methods may achieve high accuracy but they are hardly understandable by decision-makers. Models relying on symbolic, e.g. rule based methods can be less accurate but more intuitive. In both cases, feature subset selection leads to an increase of interpretability and practical usefulness of decision support systems. In this chapter, we discuss how rough sets can contribute in this respect.
In this article we present the new machine learning framework called NRough. It is focused on rou... more In this article we present the new machine learning framework called NRough. It is focused on rough set based algorithms for feature selection and classification i.e. computation of various types of decision reducts, bireducts, decision reduct ensembles and rough set inspired decision rule induction. Moreover, the framework contains other routines and algorithms for supervised and unsupervised learning. NRough is written in C# and compliant with .NET Common Language Specification (CLS). Its architecture allows easy extendability and integration.
We introduce a new rough set inspired approach to attribute selection. We consider decision syste... more We introduce a new rough set inspired approach to attribute selection. We consider decision systems with attributes specified by means of two layers: 1) general meta-attribute descriptions, and 2) their specific realizations obtained by setting up parameters of procedures calculating attribute values. We adopt methods designed for finding rough set reducts within the sets of attributes grouped into clusters, where each cluster contains potentially infinite amount of attributes realizing a single meta-attribute. As a case study, we discuss a rough set framework for multi-spectral Magnetic Resonance Image (MRI) segmentation.
We discuss several new methods for constructing approximate decision reducts from the rough set t... more We discuss several new methods for constructing approximate decision reducts from the rough set theory. We introduce generalized approximate majority decision reducts, which are an extension to standard approximate decision reducts known from literature but with improved calculation performance and complexity. We also discuss the relationship and differences of the new approximate decision reduct notion with so called decision bireducts-another type of approximate decision reducts.
A Rough Set-Based Magnetic Resonance Imaging Partial Volume Detection System
Lecture Notes in Computer Science, 2005
Application of rough set based dynamic parameter optimization to MRI segmentation
IEEE Annual Meeting of the Fuzzy Information, 2004. Processing NAFIPS '04., 2004
Springer eBooks, 2010
We propose a framework for experimental verification whether mechanisms of voting among rough-set... more We propose a framework for experimental verification whether mechanisms of voting among rough-set-based classifiers and criteria for extracting those classifiers from data should follow analogous mathematical principles. Moreover, we show that some of types of criteria perform better for high-quality data while the others are useful rather for low-quality data. The framework is based on the principles of approximate attribute reduction and probabilistic extensions of rough-set-based approach to data analysis. The framework is not supposed to produce the best-ever classification results, unless it is extended by some additional parameters known from the literature. Instead, our major goal is to illustrate in a possibly simplistic way that it is worth unifying mathematical background for the stages of learning and applying rough-set-based classifiers.
Evolutionary inspired optimization of feature subset ensembles
Rough-Set-Inspired Feature Subset Selection, Classifier Construction, and Rule Aggregation
Lecture Notes in Computer Science, 2011
ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from da... more ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from data. Each of feature subsets yields a single classifier, basically by generating its corresponding if-then decision rules from the training data. Feature subsets are extracted according to a simple randomized algorithm, following the filter (rather than wrapper or embedded) methodology. Classifier ensemble is built from single classifiers by defining aggregation laws on top of decision rules. We investigate whether rough-set-inspired methods can help in the steps of formulating feature subset optimization criteria, feature subset search heuristics, and the strategies of voting among classifiers. Comparing to our previous research, we pay a special attention to synchronization of the filter-based criteria for feature subset selection and extraction of rules basing on the obtained feature subsets. The overall framework is not supposed to produce the best-ever classification results, unless it is extended by some additional techniques known from the literature. Our major goal is to illustrate in a possibly simplistic way some general interactions between the above-mentioned criteria.
Approximation Degrees in Decision Reduct-Based MRI Segmentation
Rough-Set-based classifier extraction and voting criteria are Applied together?
Lecture Notes in Computer Science, 2010
Set Based Dynamic Parameter to MRI Segmentation
Lecture Notes in Computer Science, 2010
We propose a framework for experimental verification whether mechanisms of voting among rough-set... more We propose a framework for experimental verification whether mechanisms of voting among rough-set-based classifiers and criteria for extracting those classifiers from data should follow analogous mathematical principles. Moreover, we show that some of types of criteria perform better for high-quality data while the others are useful rather for low-quality data. The framework is based on the principles of approximate attribute reduction and probabilistic extensions of rough-set-based approach to data analysis. The framework is not supposed to produce the best-ever classification results, unless it is extended by some additional parameters known from the literature. Instead, our major goal is to illustrate in a possibly simplistic way that it is worth unifying mathematical background for the stages of learning and applying rough-set-based classifiers.
A Hybrid Approach to MR Imaging Segmentation Using Unsupervised Clustering and Approximate Reducts
Lecture Notes in Computer Science, 2005
An Automated Multi-spectral MRI Segmentation Algorithm Using Approximate Reducts
Lecture Notes in Computer Science, 2004
A Rough Set-Based Magnetic Resonance Imaging Partial Volume Detection System
Lecture Notes in Computer Science, 2005
Application of rough set based dynamic parameter optimization to MRI segmentation
IEEE Annual Meeting of the Fuzzy Information, 2004. Processing NAFIPS '04., 2004
Rough-Set-Inspired Feature Subset Selection, Classifier Construction, and Rule Aggregation
Lecture Notes in Computer Science, 2011
ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from da... more ABSTRACT We consider a rough-set-inspired framework for deriving feature subset ensembles from data. Each of feature subsets yields a single classifier, basically by generating its corresponding if-then decision rules from the training data. Feature subsets are extracted according to a simple randomized algorithm, following the filter (rather than wrapper or embedded) methodology. Classifier ensemble is built from single classifiers by defining aggregation laws on top of decision rules. We investigate whether rough-set-inspired methods can help in the steps of formulating feature subset optimization criteria, feature subset search heuristics, and the strategies of voting among classifiers. Comparing to our previous research, we pay a special attention to synchronization of the filter-based criteria for feature subset selection and extraction of rules basing on the obtained feature subsets. The overall framework is not supposed to produce the best-ever classification results, unless it is extended by some additional techniques known from the literature. Our major goal is to illustrate in a possibly simplistic way some general interactions between the above-mentioned criteria.
Approximation Degrees in Decision Reduct-Based MRI Segmentation
2007 Frontiers in the Convergence of Bioscience and Information Technologies, 2007
Evolutionary inspired optimization of feature subset ensembles
2010 Second World Congress on Nature and Biologically Inspired Computing (NaBIC), 2010
ABSTRACT
Rough Set Based Decision Support—Models Easy to Interpret
Advanced Information and Knowledge Processing, 2012
Rapid evolution of technology allows people to record more data than ever. Gathered information i... more Rapid evolution of technology allows people to record more data than ever. Gathered information is intensively used by data analysts and domain experts. Collections of patterns extracted from data compose models (compact representations of discovered knowledge), which are at the heart of each decision support system. Models based on mathematically sophisticated methods may achieve high accuracy but they are hardly understandable by decision-makers. Models relying on symbolic, e.g. rule based methods can be less accurate but more intuitive. In both cases, feature subset selection leads to an increase of interpretability and practical usefulness of decision support systems. In this chapter, we discuss how rough sets can contribute in this respect.
In this article we present the new machine learning framework called NRough. It is focused on rou... more In this article we present the new machine learning framework called NRough. It is focused on rough set based algorithms for feature selection and classification i.e. computation of various types of decision reducts, bireducts, decision reduct ensembles and rough set inspired decision rule induction. Moreover, the framework contains other routines and algorithms for supervised and unsupervised learning. NRough is written in C# and compliant with .NET Common Language Specification (CLS). Its architecture allows easy extendability and integration.
We introduce a new rough set inspired approach to attribute selection. We consider decision syste... more We introduce a new rough set inspired approach to attribute selection. We consider decision systems with attributes specified by means of two layers: 1) general meta-attribute descriptions, and 2) their specific realizations obtained by setting up parameters of procedures calculating attribute values. We adopt methods designed for finding rough set reducts within the sets of attributes grouped into clusters, where each cluster contains potentially infinite amount of attributes realizing a single meta-attribute. As a case study, we discuss a rough set framework for multi-spectral Magnetic Resonance Image (MRI) segmentation.
We discuss several new methods for constructing approximate decision reducts from the rough set t... more We discuss several new methods for constructing approximate decision reducts from the rough set theory. We introduce generalized approximate majority decision reducts, which are an extension to standard approximate decision reducts known from literature but with improved calculation performance and complexity. We also discuss the relationship and differences of the new approximate decision reduct notion with so called decision bireducts-another type of approximate decision reducts.