9 research outputs found
Non-Monotonicity in Empirical Learning Curves: Identifying non-monotonicity through slope approximations on discrete points
No full textLearning curves are used to shape the performance of a Machine Learning (ML) model with respect to the size of the set used for training it. It was commonly thought that adding more training samples would increase the model's accuracy (i.e., they are monotone), but recent works show that may not always be the case. In other words, some learners on some problems show non-monotonic behaviour. To this extent, we introduce a new method to identify non-monotonicity in empirical learning curves by approximating the curve's slope through regression around the discrete points it is defined on.This paper formalises this metric and then evaluates its accuracy through different experiments. Finally, we run the proposed metric on a subset of the extensive Learning Curve Database (LCDB) by Mohr et al. to gain better insights into the problem of non-monotonicity of learning. We found that the metric can identify non-monotonicity in learning curves well (98% experimental accuracy) and does not consider small increases due to measurement error as non-monotonicity in the curve. Finally, we have identified that non-monotonicity may be a property of some classifiers, such as Linear Discriminant Analysis. Moreover, we identified that non-monotonicity is frequently observed in datasets with faster training times.CSE3000 Research ProjectComputer Science and Engineerin
The Expanding Thread of Ungulate Browsing—A Review of Forest Ecosystem Effects and Management Approaches in Europe
No full textIn recent decades, ungulates have expanded in number and range in Europe. This review aims to analyze the impact of ungulate browsing in different forest ecosystems and identify the main driving factors and trends. In total, 155 studies were analyzed in preparing this review, across 19 European countries. In Europe, the main browsers are represented by roe deer (Capreolus capreolus L.), red deer (Cervus elaphus L.), moose (Alces alces L.), chamois (Rupicapra rupicapra L.), and fallow deer (Dama dama L.). Regarding browsing severity, they frequently exceeded 50%, meaning that over half of the saplings were browsed. Ungulate density was the main driving factor of browsing severity, with areas exhibiting high browsing pressure often having more than ten individuals per square kilometer. The type of silvicultural system used played a vital role in the severity of browsing, and trends in foraging for preferred tree species were identified. Fencing was the most common non-harmful protection method used, while hunting management was the most efficient method for controlling deer numbers and browsing intensity. Large carnivores were missing in most study areas, but in the areas where they were present, they played a significant role in creating a chain reaction of ecological impacts. Considering the significant impact of ungulate browsing on forest ecosystems, there is a pressing need for more research to comprehend and effectively mitigate the effects of deer presence comprehensively
Collaborative Solution Architecture for Developing a National Interoperability Framework in Romania
No full textInteroperability framework is a set of standards and guidelines that describe how organizations have established or will establish to interact. The framework is not static, but one that adapts to the change of standards, administrative requirements and technology. It can be adapted to the socio - economic, political, cultural, linguistic, historical and geographical purposes and to a specific context or situation. The article aims to clarify the essential concepts necessary for outlining Romanian national interoperability framework and to propose collaborative solution architecture for its development, updating and maintaining.Interoperability Framework, National Interoperability Framework, European Interoperability Strategy, Collaborative Solution Architecture
Incursiune in contractele aleatorii. Renta viagera si intretinerea - Foray into Contracts. Life Annuity and Maintanance (Romanian Version)
No full textThis monography is devoted to the life annuity contract and maintenance contract is a valuable scientific work, of real originality, for the major institutions in civil law. We can distinguish the conscientiousness and depth of research undertaken by the author, which bases her analysis and scientific findings on a serious and extensive documentation, including not only treatment, training and monographs, but also relevant articles and studies published in professional journals and a relevant law courts, covering the doctrine and jurisprudence plan the attorney activities generated in recent years by the two contracts. Associate Professor Dr. Codrin Macovei No.pg. 290maintenance contract, civil law, annuity contract
Wood Colour Variations of Quercus Species in Romania
No full textWood colour metrics are increasingly being used in wood technology and ecology studies. Researchers usually determine the colours of the wood after treatment or in different habitats. There is very little research dedicated to the problem of colour variations among one specific species harvested in different forests or regions. The main purpose of the current research is to reveal and estimate the colour variability of oak species. For this study, a total of 89 samples were taken from the heartwood of seven oak species (Quercus robur L., Quercus cerris L., Quercus rubra L., Quercus pedunculiflora K. Koch., Quercus pubescens Willd., Quercus petraea (Matt.) Liebl., and Quercus palustris Muenchh.). The CIELAB system was used for the assessment of the colour differences. To determine the colour groups and variations, K-means clustering was used. The results show that colour variations do exist. According to the cluster analysis, at least five types of oak wood can be distinguished (because in some clusters, very few samples were present) in the investigated forests. The differences are mainly observed in terms of the lightness (L*) and yellowness (b*). Redness is not a feature by which oak wood differs, but the wood can be brighter or more yellow in some of the samples. The density of the Romanian oaks in the sample does not influence the colour coordinates. The only coordinate affected is a*, but with a very small probability
The Influence of Academic Promotion Systems on the Bibliography of the Scientific Articles Published in Romanian Medical Journals
No full textIn December 2016 the Hirsch index was proposed as a promotion indexin the academic field. This uses as a way to assess a formula that includes thenumber of the articles published in the ISI Thomson Reuters indexed journals, andwhere an essential criterion is the number of quotations made by other authorsreported to the works of the assessed author. The current study aims at analyzing,in a less exhaustive manner, the degree at which the introduction of the Hirschindex score as an academic promoting criterion has changed the attitude of theresearchers towards the bibliographic references they cite in the scientific papersthey publish in the ISI indexed journals
Correction: Rita et al. Using Deep Learning and Google Street View Imagery to Assess and Improve Cyclist Safety in London. Sustainability 2023, 15, 10270
No full textRegarding the author correction request, it has been clarified that the original paper was based on a master’s project at Imperial College London [...
Local-scale mapping of tree species in a lower mountain area using Sentinel-1 and -2 multitemporal images, vegetation indices, and topographic information
No full textIntroductionMapping tree species is an important activity that provides the information necessary for sustainable forest management. Remote sensing is a effective tool that offers data at different spatial and spectral resolutions over large areas. Free and open acces Sentinel satellite imagery and Google Earth Engine, which is a powerful cloud computing platform, can be used together to map tree species.MethodsIn this study we mapped tree species at a local scale using recent Sentinel-1 (S-1) and Sentinel-2 (S-2) time-series imagery, various vegetation indices (Normalized Difference Vegetation Index - NDVI, Enhanced Vegetation Index - EVI, Green Leaf Index - GLI, and Green Normalized Difference Vegetation Index - GNDVI) and topographic features (elevation, aspect and slope). Five sets of data were used, in different combinations, together with the Random Forest classifier in order to determine seven tree species (spruce, beech, larch, fir, pine, mixed, and other broadleaves [BLs]) in the studied area.Results and discussionDataset 1 was a combination of S-2 images (bands 2, 3, 4, 5, 6, 7, 8, 8a, 11 and 12), for which an overall accuracy of 76.74% was obtained. Dataset 2 comprised S-2 images and vegetation indices, leading to an overall accuracy of 78.24%. Dataset 3 included S-2 images and topographic features, which lead to an overall accuracy of 89.51%. Dataset 4 included S-2 images, vegetation indices, and topographic features, that have determined an overall accuracy of 89.36%. Dataset 5 was composed of S-2 images, S-1 images (VV and VH polarization), vegetation indices, and topographic features that lead to an overall accuracy of 89.68%. Among the five sets of data, Dataset 3 produced the most significant increase in accuracy, of 12.77%, compared to Dataset 1. Including the vegetation indices with the S-2 images (Dataset 2) gave an accuracy increase of only 1.50%. By combining the S-1 and S-2 images, vegetation indices and topographic features (Dataset 5) there was an accuracy increase of only 0.17%, compared with the S-2 images plus topographic features combination (Dataset 3). However, the input brought by the S-1 images was apparent in the increase in classification accuracy for the mixed and other BL species that were mostly found in hilly locations. Our findings confirm the potential of S-2 images, used together with other variables, for classifying tree species at the local scale
Leaf-based characterization of intermediate forms between Cuban and Honduran mahogany
No full textThe genus Swietenia includes two mahogany species, Swietenia macrophylla (King.), commonly known as Honduran mahogany, and Swietenia mahagoni (L.) Jacq, commonly known as Cuban mahogany. There are reports of morphologically intermediate forms between Cuban and Honduran mahogany that have been localized in some Caribbean islands. The main objective of this research is to distinguish morphologically intermediate forms between the parental species S. mahagoni and S. macrophylla based on leaf morphological characters. Phenotypic data from a total of 357 mahogany trees were collected in the province of Sancti Spiritus, Cuba. Rachis length, petiole length, leaflet length, leaflet width, number of leaflets and number of nerves were evaluated. Morphological descriptor analysis revealed significant differences in the morphometric variables of the evaluated leaves, with rachis length, number of leaflets, leaflet length, petiole length, leaflet width, and number of nerves being significantly higher in S. macrophylla than in S. mahagoni. ANOVA reflected the variance between all leaf morphological parameters evaluated between the two pure species with statistically significant difference. Using the function obtained in the first DC of the two species, the individuals in the mixed stands were classified into 71 morphologically intermediate forms, 45 S. macrophylla and 64 S. mahagoni. In the second DC with the two species and the morphologically intermediate forms, Wilks’ partial lambda indicates that the variable "leaflet length" with 0.67 contributed most to the overall discrimination
