The problem considered is how to model perception and identify behavioral patterns of objects changing over time in complex dynamical systems. An approach to solving this problem has been found in the context of rough set theory and methods. Rough set theory introduced by Zdzisaw Pawlak during the early 1980s provides the foundation for the construction of classifiers, relative to what are known as temporal pattern tables. Temporal patterns can be treated as features that make it possible to approximate complex concepts. This article introduces some rough set tools for perception modeling that are developed for a system for modeling networks of classifiers. Such networks make it possible to identify behavioral patterns of objects changing over time. They are constructed using an ontology of concepts delivered by experts that engage in approximate reasoning about concepts embedded in such an ontology. We also present a method that we call a method for on-line elimination of non-relevant parts (ENP). This method was developed for on-line elimination of complex object parts that are irrelevant for identifying a given behavioral pattern. The article includes results of experiments that have been performed on data from a vehicular traffic simulator and on medical data obtained from Neonatal Intensive Care Unit in the Department of Pediatrics, Collegium Medicum, Jagiellonian University. The contribution of this article is the introduction of a network of classifiers that make it possible to identify the behavioral patterns of objects that change over time.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
We discuss medical treatment planning in the context of case-based planning, where plans (of treatment) are treated as complex decisions. A plan for a particular case is constructed from known plans for similar training examples. In order to evaluate and improve the prediction quality of complex decisions, we use a method for approximation of similarity measure between plans. The method makes it possible to transform the acquired domain knowledge about similarities of plans, expressed by medical experts in natural language, to a low level language understandable by the system. To accomplish this task, we developed a method for approximation of the ontology of concepts expressed by medical experts. We present two applications of the ontology approximation, namely, for approximation of similarity between patient histories and for approximation of compatibility of patient histories with planned therapies. Next, we use these concept approximations to define two measures on which are based two methods for (plan) therapy prediction. The article includes results of experiments with these methods performed on medical data obtained from Neonatal Intensive Care Unit, First Department of Pediatrics, Polish-American Institute of Pediatrics, Collegium Medicum, Jagiellonian University, Kraków, Poland. The experiments are pertained to the identification of infants' death risk caused by respiratory failure.
4
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
The decision making depends on the perception of the world and the proper identification of objects. The perception can be modified by various factors, that alter a way of perceiving the object even though the object is not changed (e.g., in the perception of a medical condition, such factors can be drugs or diet). The purpose of this research is to study how the disturbing factors can influence the perception. The idea was to introduce the description of the rules of these changes. We propose a method for evaluating the effect of additional therapy in patients with coronary heart disease based on the tree of the impact. The leaves of the tree provide crossdecision rules of perception changes which could be suggested as a solution to the problem of predicting changes in perception. The problems considered in this paper are associated with the design of classifiers which allow the perception of the object in the context of information related to the decision attribute.
5
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
This article introduces a new method of a decision tree construction. Such construction is performed using additional cuts applied for a verification of the cuts' quality in tree nodes during the classification of objects. The presented approach allows us to exploit the additional knowledge represented in the attributes which could be eliminated using greedy methods. The paper includes the results of experiments performed on data sets from a biomedical database and machine learning repositories. In order to evaluate the presented method, we compared its performance with the classification results of a local discretization decision tree, well known from literature. Our new method out performs the existing method, which is also confirmed by statistical tests.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.