Search results for "information extraction"
showing 5 items of 25 documents
<title>Spectral/spatial integration effects on information extraction from multispectral data: multiresolution approaches</title>
1995
New techniques for information extraction from multispectral data require physical modeling to understand the energy transfer at the atmosphere/surface interface and to develop appropriate inversion procedures, in combination with advanced processing techniques. A multi-step procedure is proposed in this work: the first step implies a binary decision about the second step to be applied in each case. If the pixel is considered as being a `pure' pixel, through a spectral/spatial classification procedure based on multiresolution techniques, then numerical inversion techniques, based on a multiple-scattering reflectance model, are used to extract parameters representing specific surface propert…
Dynamic integration of classifiers in the space of principal components
2003
Recent research has shown the integration of multiple classifiers to be one of the most important directions in machine learning and data mining. It was shown that, for an ensemble to be successful, it should consist of accurate and diverse base classifiers. However, it is also important that the integration procedure in the ensemble should properly utilize the ensemble diversity. In this paper, we present an algorithm for the dynamic integration of classifiers in the space of extracted features (FEDIC). It is based on the technique of dynamic integration, in which local accuracy estimates are calculated for each base classifier of an ensemble, in the neighborhood of a new instance to be pr…
Attention Direction in Static and Animated Diagrams
2010
Two key requirements for comprehending a diagram are to parse it into appropriate components and to establish relevant relationships between those components. These requirements can be particularly demanding when the diagram is complex and the viewers are novices in the depicted domain. Lack of domain-specific knowledge for top-down guidance of visual attention prejudices novices' extraction of task-relevant information. Static diagrams designed for novices often include visual cues intended to improve such information extraction. However, because current approaches to cueing tend to be largely intuitive, their effectiveness can be questionable. Further, animated diagrams with their percept…
<title>Expanding context against weighted voting of classifiers</title>
2000
In the paper we propose a new method to integrate the predictions of multiple classifiers for Data Mining and Machine Learning tasks. The method assumes that each classifier stands in it's own context, and the contexts are partially ordered. The order is defined by monotonous quality function that maps each context to the value from the interval [0,1]. The classifier that has the context with better quality is supposed to predict better than the classifier from worse quality. The objective is to generate the opinion of `virtual' classifier that stands in the context with quality equal to 1. This virtual classifier must have the best accuracy of predictions due to the best context. To do thi…
<title>Distance functions in dynamic integration of data mining techniques</title>
2000
One of the most important directions in the improvement of data mining and knowledge discovery is the integration of multiple data mining techniques. An integration method needs to be able either to evaluate and select the most appropriate data mining technique or to combine two or more techniques efficiently. A recent integration method for the dynamic integration of multiple data mining techniques is based on the assumption that each of the data mining techniques is the best one inside a certain subarea of the whole domain area. This method uses an instance-based learning approach to collect information about the competence areas of the mining techniques and applies a distance function to…