László Szathmáry's research while affiliated with University of Debrecen and other places

Publications (18)

Article
Full-text available
Non-Technical Loss (NTL) is a major concern for many electric supply companies due to the financial impact caused as a result of suspect consumption activities. A range of machine learning classifiers have been tested across multiple synthesized and real datasets in order to combat NTL. An important characteristic that exists in these datasets is t...
Article
Kutatócsoportunk északkelet-magyarországi régészeti ásatások során előkerült mesterségesen torzított koponyák több szempontú antropológiai vizsgálatával foglalkozik. A koponyák hun-germán korúnak datált (5–6. század) sírokból származnak. A mesterséges torzítás, mint szokás, koponyára tett hatásának minél részletesebb megismeréséhez szükséges, hogy...
Article
Full-text available
Power companies are responsible for producing and transferring the required amount of electricity from grid stations to individual households. Many countries suffer huge losses in billions of dollars due to non-technical loss (NTL) in power supply companies. To deal with NTL, many machine learning classifiers have been employed in recent time. Howe...
Article
Full-text available
With the ever-growing demand of electric power, it is quite challenging to detect and prevent Non-Technical Loss (NTL) in power industries. NTL is committed by meter bypassing, hooking from the main lines, reversing and tampering the meters. Manual on-site checking and reporting of NTL remains an unattractive strategy due to the required manpower a...
Article
Full-text available
The study’s goal is to propose a unified descriptive method for analysing artificial cranial deformation (ACD) in order to achieve consistent descriptions as well as to document the modification from a palaeopathological perspective. A number of examinations have dealt with the fact that ACD can either directly cause or predispose the formation of...
Article
Full-text available
In the present paper the authors compared skeletal populations (2421 individuals) excavated from four cemeteries, namely Hajdúdorog-Gyúlás (10th century AD), Hajdidorog-Temetöhegy (11th century AD), Hajdúdorog-Katidülö (12th-13th century AD) and Hajdúdorog-Szálldáföld (12th-13th century AD) from a micro-region of Northern Hajdúság (located in the n...
Article
Full-text available
From an anthropological point of view, artificial deformation of the cranial shape in newborns is one of the most interesting human customs, which has been recorded in all continents and in different cultures. However, the main goals of this procedure were basically the same everywhere; that is, to distinguish certain groups of people from others a...
Article
Full-text available
Researchers have dealt with the issue of the artificial cranial deformation for a long time. Cranial deformation by various devices could be found in every historical era, in all continents. It is still a custom even in present days in some places. According to the anthropological studies, multitudinous appearance of the custom of artificial crania...
Article
Full-text available
Craniometrical and anatomical investigations discovered that the structure of the populations living in the Alföld region (Hungary) might be quite different between the age of the Hungarian conquest (10th century) and the Arpadian age (11th-13th century). This statement was presumed on the basis of skeletons derived from cemeteries dating from both...
Article
Full-text available
János I., Szathmáry L., Kiss F., Hüse L.: Paleodemographical relationships in a micro-region (Northern-Hajdúság) in Tiszántúl in the X–XIIIth c.: In the present study the authors compared the populations of four cemeteries, namely Hajdúdorog-Gyúlás (Xth c.), Hajdúdorog-Temetőhegy (XIth c.), Hajdúdorog-Katidülő (XII–XIIIth c.) and Hajdúdorog-Szállás...
Article
Le sujet principal de cette thèse est l'extraction de connaissances dans les bases de données (ECBD). Plus précisément, nous avons étudié deux des plus importantes tâches d'ECBD actuelles, qui sont l'extraction de motifs et la génération de règles d'association. Tout au long de notre travail, notre objectif a été de trouver des règles d'association...

Citations

... The classifiers used in this study are support vector machine (SVM), Gradient boosting (GB) classifier, Extreme Gradient boosting (XGB) classifier, catboost (CB) classifier, AdaBoost classifier using decision tree (ADA-DT) and AdaBoost classifier using random forest (ADA-RF) [10,[29][30][31][32][33]. The choice of classifiers is influenced by the fact that some of these are preferred due to their ensemble properties of combining the weak learners and improving the performance by collective or majority learning, while others, such as SVM, are widely used due to their hyper plane properties to create significant margin, thus achieving high performance [34,35]. ...
... The classifiers used in this study are support vector machine (SVM), Gradient boosting (GB) classifier, Extreme Gradient boosting (XGB) classifier, catboost (CB) classifier, AdaBoost classifier using decision tree (ADA-DT) and AdaBoost classifier using random forest (ADA-RF) [10,[29][30][31][32][33]. The choice of classifiers is influenced by the fact that some of these are preferred due to their ensemble properties of combining the weak learners and improving the performance by collective or majority learning, while others, such as SVM, are widely used due to their hyper plane properties to create significant margin, thus achieving high performance [34,35]. ...
... Ref. [6] studied how best algorithmic bias can be detected and mitigated when applying machine-learning algorithms in the making of both simple and complex decision processes. The effectiveness of various machines learning classification models have also been assessed by various authors [9], [2], [10], [3]. Ref. [8] specifically points out that the effectiveness of a machine learning solution is directly linked to the data's characteristic and nature as well as the learning algorithm's performance. ...
... One of the machine learning algorithms used to detect the FDIs is the support vector machines (SVM). The SVM is widely used for classification problems [33,34] in a variety of different domains due to their high predicting power, high margin and the use of the support vectors to better fit the data and their capabilities of handling the data with outliers [33]. Therefore, SVM is quite efficient machine learning algorithm, easy to understand, implement and interpret. ...
... The problem faced if there is no balance between technology and the environment are air, water, and soil pollutions. High growth in urban and suburban areas has great potential in increasing the use of energy consumption, such as fuel requirements for electricity generation, industrial furnaces, and transportation [4][5][6]. ...
... Setting it arbitrarily low will qualify itemsets that should be left out, vice versa. Moreover, as database size increases, the support threshold may need to be adjusted [28] [29]. ...
... For more than three decades the Formal Concept Analysis (FCA) is being developed successfully at the intersection of applied mathematics and computer science [1][2][3][4][5][6][7]. FCA has made a significant contribution and will continue to stimulate the developing of data mining, data representation and other parts of computer science due to the classical (Aristotelian) approach to the concept as the fundamental mental entity defined by the volume and content as well as to the basis of algebraic lattices theory. ...
... In addition, certain traumas have been considered lesions resulting from riding accidents, in particular following Anđelinović et al. (2015), Khudaverdyan et al. (2016), and Ki et al. (2018). Artificial cranial deformation (ACD) has been evaluated following Dembo and Imbelloni (1938), Kiszely 1978, Tiesler 2014, and Molnár et al. (2014. ...
... hydroxyapatite provides a good approximation to bone general composition in vivo: around 380 g/kg of Ca, 180 g/kg of P with traces of Na (6 g/kg), Mg (6 g/kg) and other elements (Burton and Katzenberg, 2019). In this study, Ca content was within the range (~220-280 g/kg) of the values reported by other investigations in archaeological and modern skeletal samples (e.g., Allmäe et al., 2012;János et al., 2011;López-Costas et al., 2016;Nganvongpanit et al., 2016b;Szostek et al., 2006;Zapata et al., 2006). Phosphorus levels were also similar to those observed in osteological samples, approximately between 98 and 180 g/kg (e.g., Grupe, 1988;János et al., 2011;López-Costas et al., 2016;Miculescu et al., 2012;Zapata et al., 2006). ...
... In the preliminary version of our tool, the closed patterns are computed using the Charm algorithm [83] provided by the Coron platform [84]. Since the tool implementing our classifier is more a proof of concept than a working tool, we do not use the monotony of the measures to prune the dataspace but rather compute all the closed patterns and choose the ones that correspond to our heuristics. ...