Two types of heuristic estimators based on Parzen kernels are presented. They are able to estimate the regression function in an incremental manner. The estimators apply two techniques commonly used in concept-drifting data streams, i.e., the forgetting factor and the sliding window. The methods are applicable for models in which both the function and the noise variance change over time. Although nonparametric methods based on Parzen kernels were previously successfully applied in the literature to online regression function estimation, the problem of estimating the variance of noise was generally neglected. It is sometimes of profound interest to know the variance of the signal considered, e.g., in economics, but it can also be used for determining confidence intervals in the estimation of the regression function, as well as while evaluating the goodness of fit and in controlling the amount of smoothing. The present paper addresses this issue. Specifically, variance estimators are proposed which are able to deal with concept drifting data by applying a sliding window and a forgetting factor, respectively. A number of conducted numerical experiments proved that the proposed methods perform satisfactorily well in estimating both the regression function and the variance of the noise.
The aim of this research was to find out what changes occurred between 1999-2009 in Oszast reserve in the volume, species composition and diameter at breast height (dbh) distribution of the forest stand, and the number and height of regeneration. The objective was to determine what would be condition of these managed lower montane multispecies forest stands (Swiss irregular shelterwood method or selection cuttings) and what role spruce would play in them if they have not been replaced by spruce monocultures. The research was conducted on three permanent circular sample plots (s.p.), each had size of 1/3 ha. Over 10 years, standing volume of the forest stand increased on s.p. 1 (from around 562 m3/ha to 649 m3/ha) and s.p. 3. (from 653 m3/ha to 660 m3/ha), while decreased on s.p. 2. (from 421 m3/ha to 378 m3/ha). The species composition, defined on the basis of volume share (averaged for the three s.p. jointly), did not undergo consistent changes. However, the relative dominance of beech over spruce was determined based on tree numbers. The average spruce mortality (averaged from three s.p.) did not exceed 10% and was slightly higher than that of beech (6%), and lower than fir mortality (15%). Nevertheless, spruce did not show any symptoms of dieback. The reasons behind its mortality were fallen trees and windbreaks. In regeneration, on the whole, beech or sycamore predominated, and the proportion of spruce and fir was small. In the future spruce and fir may even decrease further by competitive ability of dynamically regenerating beech. Abandonment of forest management to promote greater diversity of species, may favour the formation of beech monocultures, or forest stands dominated by beech, everywhere that beech is already present or will be introduced. The maintenance of stable, multispecies forest stands, with co-dominant fir, beech and spruce of native origin, requires natural or artificial regeneration of spruce and fir, manipulated to restore fir up to about 30%, and reduce spruce down to about 40%. This would be possible through the use of the Swiss irregular shelterwood method and selection system, and by continuous tending of regeneration
Introduction: The aim of the study was to analyse selected densitometric and geometric parameters in the third metacarpal bone along the long axis in horses. The densitometric parameters included the cortical and trabecular bone mineral density, while the geometric parameters included the cortical, trabecular, and total areas, strength strain index X, strength strain index Y, and the polar strength strain index.
Material and Methods: The parameters were analysed using eight sections from 10% to 80% of the length of the bone. Peripheral quantitative computed tomography was used in the study. Statistical analysis was carried out using the Friedman analysis of variance and post-hoc tests.
Results: The proximal metaphyseal region showed the highest predicted resistance to bone fractures in the transverse (back-front) plane, the distal metaphyseal region had the highest predicted resistance to transverse and torsional fractures in the transverse (side-side) plane. The cross-sectional area and the shape of the cross-section of the cortical bone of the MCIII had the highest coefficient of variation. The density of the cortical bone was least variable.
Conclusions: The cortical area and cortical bone mineral density assumed the highest values in the diaphyseal region, while the highest total area, trabecular area and trabecular bone mineral density values were obtained in the metaphyseal proximal and distal region.
In recent years, many deep learning methods, allowed for a significant improvement of systems based on artificial intelligence methods. Their effectiveness results from an ability to analyze large labeled datasets. The price for such high accuracy is the long training time, necessary to process such large amounts of data. On the other hand, along with the increase in the number of collected data, the field of data stream analysis was developed. It enables to process data immediately, with no need to store them. In this work, we decided to take advantage of the benefits of data streaming in order to accelerate the training of deep neural networks. The work includes an analysis of two approaches to network learning, presented on the background of traditional stochastic and batch-based methods.