In the rapidly evolving landscape of television advertising,optimizing ad schedules to maximize viewer engagement and revenue has become significant.Traditional methods often operate in silos,limiting the potential in...In the rapidly evolving landscape of television advertising,optimizing ad schedules to maximize viewer engagement and revenue has become significant.Traditional methods often operate in silos,limiting the potential insights gained from broader data analysis due to concerns over privacy and data sharing.This article introduces a novel approach that leverages Federated Learning(FL)to enhance TV ad schedule optimization,combining the strengths of local optimization techniques with the power of global Machine Learning(ML)models to uncover actionable insights without compromising data privacy.It combines linear programming for initial ads scheduling optimization with ML—specifically,a K-Nearest Neighbors(KNN)model—for predicting ad spot positions.Taking into account the diversity and the difficulty of the ad-scheduling problem,we propose a prescriptivepredictive approach in which first the position of the ads is optimized(using Google’s OR-Tools CP-SAT)and then the scheduled position of all ads will be the result of the optimization problem.Second,this output becomes the target of a predictive task that predicts the position of new entries based on their characteristics ensuring the implementation of the scheduling at large scale(using KNN,Light Gradient Boosting Machine and Random Forest).Furthermore,we explore the integration of FL to enhance predictive accuracy and strategic insight across different broadcasting networks while preserving data privacy.The FL approach resulted in 8750 ads being precisely matched to their optimal category placements,showcasing an alignment with the intended diversity objectives.Additionally,there was a minimal deviation observed,with 1133 ads positioned within a one-category variance from their ideal placement in the original dataset.展开更多
The paper addresses the impacts of political exits on capital structure policies,through studying the impact of the Brexit deal on the UK's companies listed at FTSE 100 as a practical case.The study is divided int...The paper addresses the impacts of political exits on capital structure policies,through studying the impact of the Brexit deal on the UK's companies listed at FTSE 100 as a practical case.The study is divided into two periods:the first before the Brexit deal(2010-2015)and the second after the referendum until Brexit(2016-2021).Data were collected from the annual reports of the UK companies and the FTSE 100 database.Moreover,capital structure policies were studied by analyzing the determinants such as long-term debt over total equity,profitability,tangibility,liquidity,propensity to pay dividends,size,and political factor-Brexit.The results show that there is a direct relationship between political exits and capital structure.Also,the results show that the financial ratios are different between the two periods(before and after Brexit),where it seems that the capital structure variables were affected starting from the referendum period,political negotiations between the two parties,and then Brexit(2016-2021).In addition,the correlation matrix shows that long-term debt over total equity has a negative correlation with profitability,liquidity,and political factors,where it seems that long-term debt over total equity has a positive correlation with tangibility,size,and paying dividends.Moreover,the results show that political shocks were affecting the FTSE 100 performance starting from 2016 until 2021.The study could be useful for financial and economic academics.Moreover,the study could be helpful for EU members and the UK government,taking into consideration the interests of the UK and EU where any future policies or laws will affect the financial and economic sectors of the two parties.展开更多
Extracting data from visually rich documents and charts using traditional methods that rely on OCR-based parsing poses multiple challenges,including layout complexity in unstructured formats,limitations in recognizing...Extracting data from visually rich documents and charts using traditional methods that rely on OCR-based parsing poses multiple challenges,including layout complexity in unstructured formats,limitations in recognizing visual elements,and the correlation between different parts of the documents,as well as domain-specific semantics.Simply extracting text is not sufficient;advanced reasoning capabilities are proving to be essential to analyze content and answer questions accurately.This paper aims to evaluate the ability of the Large Language Models(LLMs)to correctly answer questions about various types of charts,comparing their performance when using images as input versus directly parsing PDF files.To retrieve the images from the PDF,ColPali,a model leveraging state-of-the-art visual languagemodels,is used to identify the relevant page containing the appropriate chart for each question.Google’s Gemini multimodal models were used to answer a set of questions through two approaches:1)processing images derived from PDF documents and 2)directly utilizing the content of the same PDFs.Our findings underscore the limitations of traditional OCR-based approaches in visual document understanding(VrDU)and demonstrate the advantages of multimodal methods in both data extraction and reasoning tasks.Through structured benchmarking of chart question answering(CQA)across input formats,our work contributes to the advancement of chart understanding(CU)and the broader field of multimodal document analysis.Using two diverse and information-rich sources:the World Health Statistics 2024 report by theWorld Health Organisation and the Global Banking Annual Review 2024 by McKinsey&Company,we examine the performance ofmultimodal LLMs across different input modalities,comparing their effectiveness in processing charts as images versus parsing directly from PDF content.These documents were selected due to their multimodal nature,combining dense textual analysis with varied visual representations,thus presenting realistic challenges for vision-language models.This comparison is aimed at assessing how advanced models perform with different input formats and to determine if an image-based approach enhances chart comprehension in terms of accurate data extraction and reasoning capabilities.展开更多
The potential of text analytics is revealed by Machine Learning(ML)and Natural Language Processing(NLP)techniques.In this paper,we propose an NLP framework that is applied to multiple datasets to detect malicious Unif...The potential of text analytics is revealed by Machine Learning(ML)and Natural Language Processing(NLP)techniques.In this paper,we propose an NLP framework that is applied to multiple datasets to detect malicious Uniform Resource Locators(URLs).Three categories of features,both ML and Deep Learning(DL)algorithms and a ranking schema are included in the proposed framework.We apply frequency and prediction-based embeddings,such as hash vectorizer,Term Frequency-Inverse Dense Frequency(TF-IDF)and predictors,word to vector-word2vec(continuous bag of words,skip-gram)from Google,to extract features from text.Further,we apply more state-of-the-art methods to create vectorized features,such as GloVe.Additionally,feature engineering that is specific to URL structure is deployed to detect scams and other threats.For framework assessment,four ranking indicators are weighted:computational time and performance as accuracy,F1 score and type error II.For the computational time,we propose a new metric-Feature Building Time(FBT)as the cutting-edge feature builders(like doc2vec or GloVe)require more time.By applying the proposed assessment step,the skip-gram algorithm of word2vec surpasses other feature builders in performance.Additionally,eXtreme Gradient Boost(XGB)outperforms other classifiers.With this setup,we attain an accuracy of 99.5%and an F1 score of 0.99.展开更多
International portfolio management is influenced by the existence of“frictions”,factors or events that interfere with trade,which are linked in financial literature to market-specific factors,such as available infor...International portfolio management is influenced by the existence of“frictions”,factors or events that interfere with trade,which are linked in financial literature to market-specific factors,such as available information,restrictions,investor protection,or market liquidity.Given the wide variety of factors that can be included in these categories,scientific studies typically focus on a reduced number of indicators at a time in order to offer an in depth analysis of their impact.We offer a consolidated view of the perspectives observed in financial literature by proposing a novel index for market frictions that includes all these four components and rank fifteen post-communist East European capital markets based on their index values.We then constructed various scenarios by assuming different levels of importance for the criteria used in index construction.By employing grey clustering analysis,we cluster these capital markets into three categories—strongly recommended,recommended with some reserve,and not recommended—based on the importance given by the decision maker to these factors.The results show that some of the studied markets are in the same cluster,irrespective of the chosen scenario.The only market always included in the“strongly recommended”category is Hungary,indicating that it is a good investment option for international participants.Bulgaria and Slovakia are always regarded as“recommended with reserve”markets,whereas the Republic of Moldova is part of the“not recommended”category.The other markets show a degree of variability that can be explained by different investor perspectives.This study contributes to the existing literature by combining the advantages of grey clustering and portfolio analysis.Investors can use this approach during the decision-making process related to their investments.展开更多
In this article, we consider the existence of positive radial solutions for Hessian equations and systems with weights and we give a necessary condition as well as a sufficient condition for a positive radial solution...In this article, we consider the existence of positive radial solutions for Hessian equations and systems with weights and we give a necessary condition as well as a sufficient condition for a positive radial solution to be large. The method of proving theorems is essentially based on a successive approximation technique. Our results complete and improve a work published recently by Zhang and Zhou（existence of entire positive k-convex radial solutions to Hessian equations and systems with weights. Applied Mathematics Letters,Volume 50, December 2015, Pages 48–55）.展开更多
The essence of this study is to investigate the influence of the board gender diversity on firms’accounting and market-based performance using a sample of Standard&Poor’s 500 companies belonging to the informati...The essence of this study is to investigate the influence of the board gender diversity on firms’accounting and market-based performance using a sample of Standard&Poor’s 500 companies belonging to the information technology sector over 12 years.Using the pooled ordinary least squares(OLS)method,the outcomes provide evidence for a positive influence of women on corporate boards on both measures of company performance,except for the percentage of female executives in the case of return on assets(ROA).After estimating the fixed effects and random-effects through panel data,the econometric outcomes show no statistically significant association among board gender diversity and ROA but a positive influence of the number and percentage of women on board on price-to-earnings ratio.展开更多
Viruses can alter the expression of host microRNAs(miRNA s) and modulate the immune response during a persistent infection. The dysregulation of host miRNA s by hepatitis B virus(HBV) contributes to the proinflammator...Viruses can alter the expression of host microRNAs(miRNA s) and modulate the immune response during a persistent infection. The dysregulation of host miRNA s by hepatitis B virus(HBV) contributes to the proinflammatory and profibrotic changes within the liver. Multiple studies have documented the differential regulation of intracellular and circulating miRNA s during different stages of HBV infection. Circulating miRNA s found in plasma and/or extracellular vesicles can integrate data on viral-host interactions and on the associated liver injury. Hence, the detection of circulating miRNA s in chronic HBV hepatitis could offer a promising alternative to liver biopsy, as their expression is associated with HBV replication, the progression of liver fibrosis,and the outcome of antiviral treatment. The current review explores the available data on miRNA involvement in HBV pathogenesis with an emphasis on their potential use as biomarkers for liver fibrosis.展开更多
The term “globalization” designates the rapidly advancing international economic integration through substantial growth in trade of goods and services as well as surging cross-border factor mobility from the early 1...The term “globalization” designates the rapidly advancing international economic integration through substantial growth in trade of goods and services as well as surging cross-border factor mobility from the early 1990s onwards. This acceleration in the liberalization of world trade and capital movements is largely attributable to technological progress which significantly curtailed the expenses for transport as well as communication. In conjunction with the progressing opening of major newly industrializing economies endowed with abundant and cheap labor forces, it has reinforced a geographical fragmentation of production processes according to cost considerations. This phenomenon termed as “vertical specialization”, “outsourcing” or also “slicing the value chain” in turn has translated into a hike in FDI and international trade of intermediate products. The main objective of this paper is to identify adequate trade models for assessing the corresponding effects of those developments. In this context, textbook trade theories spanning from comparative advantage à la Ricardo to the new-new trade theory are examined. Subsequently, we expand on globalization-induced new forms of trade （intra-firm trade, trade within the value chain） and finally conclude with our findings to properly account for those trends.展开更多
The basement of the Romanian Carpathians is made of Neoproterozoic to early Paleozoic peri- Gondwanan terranes variably involved in the Variscan orogeny, similarly to other basement terrains of Europe. They were hardl...The basement of the Romanian Carpathians is made of Neoproterozoic to early Paleozoic peri- Gondwanan terranes variably involved in the Variscan orogeny, similarly to other basement terrains of Europe. They were hardly dismembered during the Alpine orogeny and traditionally have their own names in the three Carpathian areas. The Danubian domain of the South Carpathians comprises the Dragsan and Lainici-Paiius peri-Amazonian terranes. The Dragsan terrane originated within the ocean surrounding Rodinia and docked with Rodinia at -800 Ma. It does not contain Cadomian magmatism and consequently it is classified as an Avalonian extra-Cadomian terrane, The Lainici-Pfiius terrane is a Ganderian fragment strongly modified by Cadomian subduction-related magmatism, It is attached to the Moesia platform. The TisoviD terrane is an ophiolite that marks the boundary between Drfagsan and Lainici-Paius terranes. The other basement terranes of the Romanian Carpathians originated close to the Ordovician North- African orogen, as a result of the eastern Rheic Ocean opening and closure. Except for the Sebes-Lotru terrane that includes a lower metamorphic unit of Cadomian age, all the other terranes （Bretila, Tulghes, Negrisoara and Rebra in the East Carpathians, Somes, Biharia and Baia de Aries in the Apuseni mountains, Fagaras, Leaota, Carat and Pades in the South Carpathians） represent late Cambrian-Ordovician rock assemblages. Their provenance, is probably within paleo-nortbeast Africa, close to the Arabian-Nubian shield. The late Cambrian-Ordovician terranes are defined here as Carpathian-type terranes. According to their lithostratigraphy and origin, some are of continental margin magmatic arc setting, whereas others formed in rift and back-arc environment and closed to passive continental margin settings. In a paleo- geographic reconstruction, the continental margin magmatic arc terranes were first that drifted out, followed by the passive continental margin terranes with the back-arc terranes in their front. They accreted to Laurussia during the Variscan orogeny. Some of them （Sebes-Lotru in South Carpathians and Baia de Aries in Apuseni mountains） underwent eclogite-grade metamorphism. The Danubian terranes, the Bretila terrane and the Somes terrane were intruded by Variscan granitoids.展开更多
This paper explores the asymmetric effect of COVID-19 pandemic news,as measured by the coronavirus indices(Panic,Hype,Fake News,Sentiment,Infodemic,and Media Coverage),on the cryptocurrency market.Using daily data fro...This paper explores the asymmetric effect of COVID-19 pandemic news,as measured by the coronavirus indices(Panic,Hype,Fake News,Sentiment,Infodemic,and Media Coverage),on the cryptocurrency market.Using daily data from January 2020 to September 2021 and the exponential generalized autoregressive conditional heter-oskedasticity model,the results revealed that both adverse and optimistic news had the same effect on Bitcoin returns,indicating fear of missing out behavior does not prevail.Furthermore,when the nonlinear autoregressive distributed lag model is esti-mated,both positive and negative shocks in pandemic indices promote Bitcoin’s daily changes;thus,Bitcoin is resistant to the SARS-CoV-2 pandemic crisis and may serve as a hedge during market turmoil.The analysis of frequency domain causality supports a unidirectional causality running from the Coronavirus Fake News Index and Sentiment Index to Bitcoin returns,whereas daily fluctuations in the Bitcoin price Granger affect the Coronavirus Panic Index and the Hype Index.These findings may have significant policy implications for investors and governments because they highlight the impor-tance of news during turbulent times.The empirical results indicate that pandemic news could significantly influence Bitcoin’s price.展开更多
The objective of this work is the study of social and economic inequality in the space of Central and Eastern Europe and its impact on economic growth. Our study includes a three-stage methodology:(1) application of a...The objective of this work is the study of social and economic inequality in the space of Central and Eastern Europe and its impact on economic growth. Our study includes a three-stage methodology:(1) application of a clustering method based on neural network (Self Organising Maps), to the series of panel data in order to divide countries into clusters, corresponding to the degree of economic and social inequality;(2) computing a composed index of economic and social inequality, using Principal Component Analysis and an extension of the method provided by OECD for computing composite indicators;(3) constructing an econometric model to establish the impact of social and economic inequality on economic growth and a VAR model to determine the causality between main determinants to growth and inequality as well as the response to shocks to the dynamics of the variables. The 24 Eastern and Central European countries have been grouped in five clusters, according to 11 attributes. In the results obtained, the third cluster comprises countries with the most equitable income distribution: Czech Republic, Croatia, Hungary, Slovak Republic, Slovenia. To the opposite side is the fifth cluster with the deepest inequality, including only one country, namely Georgia. The second and third steps of our methodology, were applied only for the extreme clusters namely, the clusters with the highest (C5) and lowest (C3) inequality respectively.展开更多
We discussed recent studies, within the framework of transport theories, on heavy ion reactions between charge asymmetric systems, from low up to Fermi energies. We concentrated on the analysis of ternary breakup even...We discussed recent studies, within the framework of transport theories, on heavy ion reactions between charge asymmetric systems, from low up to Fermi energies. We concentrated on the analysis of ternary breakup events of dynamical origin occurring in semi-central reactions, where the formation of excited systems in various conditions of shape, excitation energy and spin is observed. At beam energies around 20 A Me V, we showed how this fragmentation mode emerges from the combined action of surface(neck) instabilities and angular momentum effects, leading to the observation of three aligned massive fragments in the exit channel. At Fermi energies, a transition towards a prompt emission of small fragments from the neck region with larger relative velocity with respect to projectile and target remnants is observed. We also focus on isospin sensitive observables, aiming at extracting information on the density dependence of the isovector part of the nuclear effective interaction and of the nuclear symmetry energy.展开更多
Using data from the Bucharest Stock Exchange,we examine the factors influencing the probability of informed trading(PIN)during February—October 2020,a COVID-19 pandemic period.Based on an unconditional quantile regre...Using data from the Bucharest Stock Exchange,we examine the factors influencing the probability of informed trading(PIN)during February—October 2020,a COVID-19 pandemic period.Based on an unconditional quantile regression approach,we show that PIN exhibit asymmetric dependency with liquidity and trading costs.Furthermore,building a customized database that contains all insider transactions on the Bucharest Stock Exchange,we reveal that these types of orders monotonically increase the infor-mation asymmetry from the 50th to the 90th quantile throughout the PIN distribution.Finally,we bring strong empirical evidence associating the level of information asym-metry to the level of fake news related to the COVID-19 pandemic.This novel result suggests that during episodes when the level of PIN is medium to high(between 15 and 50%),any COVID-19 related news classified as misinformation released during the lockdown period,is discouraging informed traders to place buy or sell orders condi-tioned by their private information.展开更多
The present study will examine the legal regime of authorized natural persons,individual enterprises,and family businesses as regulated in Romania.Such an analysis aims at presenting a panoramic view of the legal fram...The present study will examine the legal regime of authorized natural persons,individual enterprises,and family businesses as regulated in Romania.Such an analysis aims at presenting a panoramic view of the legal framework in a representative country in Eastern Europe.Business activities can be carried out individually by individuals as well as large corporations,firms,and companies that have the status of legal persons.The business carried out by individuals,mainly in terms of the quality of an authorized natural person,is particularly important because it contributes to the development of the middle class within society and to giving professional and financial independence to different categories of professionals.The main aspects related to the qualification of specific notions in the field,the administrative steps for acquiring the status of an authorized natural person,individual enterprise and family enterprise,and their legal regime will be found in the article.展开更多
The issue of this study continues the analysis of persuasive truth （the ＂truth＂ of beliefs）, a problem debated by us inseveral articles already published, especially in two of them： Doxastic Dialectics （1999） a...The issue of this study continues the analysis of persuasive truth （the ＂truth＂ of beliefs）, a problem debated by us inseveral articles already published, especially in two of them： Doxastic Dialectics （1999） and The Probable and theProblem （2010）. In this new ＂chapter,＂ our intention is to develop more details from the perspective of subjectivity,which has a grounding role in doxastic dialectics. Doxa＇s axiomatic mechanism tries to temper the subjectivedimension of persuasive truth, by submitting the doxastic proofs to the control of the oppositional principle（=antithetic ＂reasoning＂）. Doxastic thinking discovers dialectically its own ratio （=measure）, progressivelyincreasing the relevance of the listening capacity. In the text that follows, the concept of listening is used in a largerthan sensitive sense, being equivalent to condition of receptivity.展开更多
The paper aims to analyze the relationship between economic growth and increasing agricultural productivity at macroeconomic level, relationship that determines a decrease of poverty. Data sets are analyzed using econ...The paper aims to analyze the relationship between economic growth and increasing agricultural productivity at macroeconomic level, relationship that determines a decrease of poverty. Data sets are analyzed using econometric methods in order to test the cointegration relationships. The results of the data show that, in Romania’s case, the added value of agriculture has no significant impact on the evolution of the country’s GDP, so also the impact on raising the standard of living is marginal. To increase productivity and profitability of agriculture requires a complex approach: a financial system to support development of the sector by providing specific financial products and services, the reorganization/retechnologization of farms, encouraging young people to invest in agriculture, ownership structure of agricultural land.展开更多
Eye tracking can facilitate understanding irrational decision-making in contexts such as financial risk-taking.For this purpose,we develop an experimental framework in which participants trade a risky asset in a simul...Eye tracking can facilitate understanding irrational decision-making in contexts such as financial risk-taking.For this purpose,we develop an experimental framework in which participants trade a risky asset in a simulated bubble market to maximize individual returns while their eye movements are recorded.Returns are sensitive to eye movement dynamics,depending on the presented visual stimuli.Using eye-tracking data,we investigated the effects of arousal,attention,and disengagement on individual payoffs using linear and nonlinear approaches.By estimating a nonlinear model using attention as a threshold variable,our results suggest that arousal positively influences trading returns,but its effect becomes smaller when attention exceeds a certain threshold,whereas disengagement has a higher negative impact on reduced attention levels and becomes almost irrelevant when attention increases.Hence,we provide a neurobehavioral metric as a function of attention that predicts financial gains in boomand-bust scenarios.This study serves as a proof-of-concept for developing future psychometric measures to enhance decision-making.展开更多
The research paper European Tax Models is a comparative member states, taking into account the main features in light analysis of the taxation systems in the European Union of the contribution of indirect and direct t...The research paper European Tax Models is a comparative member states, taking into account the main features in light analysis of the taxation systems in the European Union of the contribution of indirect and direct taxes and social contributions to the achievement of public revenues. Theme presents a topic of great interest, both theoretically and practically, given that how to place taxes has direct repercussions on the economic development of a country, and undoubtedly influence the rules of an economy, particularly in terms of investment, labor market, and social welfare. It was considered necessary in the first part of the paper to address the conceptual elements and present the most important features of tax systems and the principles that underlie them. It was studied from a theoretical perspective and it found the European tax models as follow： Nordic, continental, Anglo-Saxon, Mediterranean, and catching-up. Then, it analyzed each fiscal European model on each member country, starting from its economic indicators, based on Eurostat data. The objective of the research paper was to present a complete picture of the structure and trends of tax level of the member states of the European Union, sorted by European tax models and the impact of taxation on economic growth and social welfare. The research has concluded that, as long as the rules of the European Union, member states are free to choose their own tax system along with their fiscal policy for economic development and having in a view of their geographical, historical, and political situation.展开更多
基金supported by a grant of the Ministry of Research,Innovation and Digitization,CNCS/CCCDI-UEFISCDI,project number COFUND-DUT-OPEN4CEC-1,within PNCDI IV.
文摘In the rapidly evolving landscape of television advertising,optimizing ad schedules to maximize viewer engagement and revenue has become significant.Traditional methods often operate in silos,limiting the potential insights gained from broader data analysis due to concerns over privacy and data sharing.This article introduces a novel approach that leverages Federated Learning(FL)to enhance TV ad schedule optimization,combining the strengths of local optimization techniques with the power of global Machine Learning(ML)models to uncover actionable insights without compromising data privacy.It combines linear programming for initial ads scheduling optimization with ML—specifically,a K-Nearest Neighbors(KNN)model—for predicting ad spot positions.Taking into account the diversity and the difficulty of the ad-scheduling problem,we propose a prescriptivepredictive approach in which first the position of the ads is optimized(using Google’s OR-Tools CP-SAT)and then the scheduled position of all ads will be the result of the optimization problem.Second,this output becomes the target of a predictive task that predicts the position of new entries based on their characteristics ensuring the implementation of the scheduling at large scale(using KNN,Light Gradient Boosting Machine and Random Forest).Furthermore,we explore the integration of FL to enhance predictive accuracy and strategic insight across different broadcasting networks while preserving data privacy.The FL approach resulted in 8750 ads being precisely matched to their optimal category placements,showcasing an alignment with the intended diversity objectives.Additionally,there was a minimal deviation observed,with 1133 ads positioned within a one-category variance from their ideal placement in the original dataset.
文摘The paper addresses the impacts of political exits on capital structure policies,through studying the impact of the Brexit deal on the UK's companies listed at FTSE 100 as a practical case.The study is divided into two periods:the first before the Brexit deal(2010-2015)and the second after the referendum until Brexit(2016-2021).Data were collected from the annual reports of the UK companies and the FTSE 100 database.Moreover,capital structure policies were studied by analyzing the determinants such as long-term debt over total equity,profitability,tangibility,liquidity,propensity to pay dividends,size,and political factor-Brexit.The results show that there is a direct relationship between political exits and capital structure.Also,the results show that the financial ratios are different between the two periods(before and after Brexit),where it seems that the capital structure variables were affected starting from the referendum period,political negotiations between the two parties,and then Brexit(2016-2021).In addition,the correlation matrix shows that long-term debt over total equity has a negative correlation with profitability,liquidity,and political factors,where it seems that long-term debt over total equity has a positive correlation with tangibility,size,and paying dividends.Moreover,the results show that political shocks were affecting the FTSE 100 performance starting from 2016 until 2021.The study could be useful for financial and economic academics.Moreover,the study could be helpful for EU members and the UK government,taking into consideration the interests of the UK and EU where any future policies or laws will affect the financial and economic sectors of the two parties.
基金supported by a grant from the Ministry of Research,Innovation and Digitization,CNCS/CCCDI-UEFISCDI,project number COFUND-CETP-SMART-LEM-1,within PNCDI Ⅳ.
文摘Extracting data from visually rich documents and charts using traditional methods that rely on OCR-based parsing poses multiple challenges,including layout complexity in unstructured formats,limitations in recognizing visual elements,and the correlation between different parts of the documents,as well as domain-specific semantics.Simply extracting text is not sufficient;advanced reasoning capabilities are proving to be essential to analyze content and answer questions accurately.This paper aims to evaluate the ability of the Large Language Models(LLMs)to correctly answer questions about various types of charts,comparing their performance when using images as input versus directly parsing PDF files.To retrieve the images from the PDF,ColPali,a model leveraging state-of-the-art visual languagemodels,is used to identify the relevant page containing the appropriate chart for each question.Google’s Gemini multimodal models were used to answer a set of questions through two approaches:1)processing images derived from PDF documents and 2)directly utilizing the content of the same PDFs.Our findings underscore the limitations of traditional OCR-based approaches in visual document understanding(VrDU)and demonstrate the advantages of multimodal methods in both data extraction and reasoning tasks.Through structured benchmarking of chart question answering(CQA)across input formats,our work contributes to the advancement of chart understanding(CU)and the broader field of multimodal document analysis.Using two diverse and information-rich sources:the World Health Statistics 2024 report by theWorld Health Organisation and the Global Banking Annual Review 2024 by McKinsey&Company,we examine the performance ofmultimodal LLMs across different input modalities,comparing their effectiveness in processing charts as images versus parsing directly from PDF content.These documents were selected due to their multimodal nature,combining dense textual analysis with varied visual representations,thus presenting realistic challenges for vision-language models.This comparison is aimed at assessing how advanced models perform with different input formats and to determine if an image-based approach enhances chart comprehension in terms of accurate data extraction and reasoning capabilities.
基金supported by a grant of the Ministry of Research,Innovation and Digitization,CNCS-UEFISCDI,Project Number PN-Ⅲ-P4-PCE-2021-0334,within PNCDI Ⅲ.
文摘The potential of text analytics is revealed by Machine Learning(ML)and Natural Language Processing(NLP)techniques.In this paper,we propose an NLP framework that is applied to multiple datasets to detect malicious Uniform Resource Locators(URLs).Three categories of features,both ML and Deep Learning(DL)algorithms and a ranking schema are included in the proposed framework.We apply frequency and prediction-based embeddings,such as hash vectorizer,Term Frequency-Inverse Dense Frequency(TF-IDF)and predictors,word to vector-word2vec(continuous bag of words,skip-gram)from Google,to extract features from text.Further,we apply more state-of-the-art methods to create vectorized features,such as GloVe.Additionally,feature engineering that is specific to URL structure is deployed to detect scams and other threats.For framework assessment,four ranking indicators are weighted:computational time and performance as accuracy,F1 score and type error II.For the computational time,we propose a new metric-Feature Building Time(FBT)as the cutting-edge feature builders(like doc2vec or GloVe)require more time.By applying the proposed assessment step,the skip-gram algorithm of word2vec surpasses other feature builders in performance.Additionally,eXtreme Gradient Boost(XGB)outperforms other classifiers.With this setup,we attain an accuracy of 99.5%and an F1 score of 0.99.
文摘International portfolio management is influenced by the existence of“frictions”,factors or events that interfere with trade,which are linked in financial literature to market-specific factors,such as available information,restrictions,investor protection,or market liquidity.Given the wide variety of factors that can be included in these categories,scientific studies typically focus on a reduced number of indicators at a time in order to offer an in depth analysis of their impact.We offer a consolidated view of the perspectives observed in financial literature by proposing a novel index for market frictions that includes all these four components and rank fifteen post-communist East European capital markets based on their index values.We then constructed various scenarios by assuming different levels of importance for the criteria used in index construction.By employing grey clustering analysis,we cluster these capital markets into three categories—strongly recommended,recommended with some reserve,and not recommended—based on the importance given by the decision maker to these factors.The results show that some of the studied markets are in the same cluster,irrespective of the chosen scenario.The only market always included in the“strongly recommended”category is Hungary,indicating that it is a good investment option for international participants.Bulgaria and Slovakia are always regarded as“recommended with reserve”markets,whereas the Republic of Moldova is part of the“not recommended”category.The other markets show a degree of variability that can be explained by different investor perspectives.This study contributes to the existing literature by combining the advantages of grey clustering and portfolio analysis.Investors can use this approach during the decision-making process related to their investments.
文摘In this article, we consider the existence of positive radial solutions for Hessian equations and systems with weights and we give a necessary condition as well as a sufficient condition for a positive radial solution to be large. The method of proving theorems is essentially based on a successive approximation technique. Our results complete and improve a work published recently by Zhang and Zhou（existence of entire positive k-convex radial solutions to Hessian equations and systems with weights. Applied Mathematics Letters,Volume 50, December 2015, Pages 48–55）.
文摘The essence of this study is to investigate the influence of the board gender diversity on firms’accounting and market-based performance using a sample of Standard&Poor’s 500 companies belonging to the information technology sector over 12 years.Using the pooled ordinary least squares(OLS)method,the outcomes provide evidence for a positive influence of women on corporate boards on both measures of company performance,except for the percentage of female executives in the case of return on assets(ROA).After estimating the fixed effects and random-effects through panel data,the econometric outcomes show no statistically significant association among board gender diversity and ROA but a positive influence of the number and percentage of women on board on price-to-earnings ratio.
基金Ministerul Cercetarii si Inovarii,Programul 1,subprogramul 1.2.Performanta institutionala,No.PFE_23/2018。
文摘Viruses can alter the expression of host microRNAs(miRNA s) and modulate the immune response during a persistent infection. The dysregulation of host miRNA s by hepatitis B virus(HBV) contributes to the proinflammatory and profibrotic changes within the liver. Multiple studies have documented the differential regulation of intracellular and circulating miRNA s during different stages of HBV infection. Circulating miRNA s found in plasma and/or extracellular vesicles can integrate data on viral-host interactions and on the associated liver injury. Hence, the detection of circulating miRNA s in chronic HBV hepatitis could offer a promising alternative to liver biopsy, as their expression is associated with HBV replication, the progression of liver fibrosis,and the outcome of antiviral treatment. The current review explores the available data on miRNA involvement in HBV pathogenesis with an emphasis on their potential use as biomarkers for liver fibrosis.
文摘The term “globalization” designates the rapidly advancing international economic integration through substantial growth in trade of goods and services as well as surging cross-border factor mobility from the early 1990s onwards. This acceleration in the liberalization of world trade and capital movements is largely attributable to technological progress which significantly curtailed the expenses for transport as well as communication. In conjunction with the progressing opening of major newly industrializing economies endowed with abundant and cheap labor forces, it has reinforced a geographical fragmentation of production processes according to cost considerations. This phenomenon termed as “vertical specialization”, “outsourcing” or also “slicing the value chain” in turn has translated into a hike in FDI and international trade of intermediate products. The main objective of this paper is to identify adequate trade models for assessing the corresponding effects of those developments. In this context, textbook trade theories spanning from comparative advantage à la Ricardo to the new-new trade theory are examined. Subsequently, we expand on globalization-induced new forms of trade （intra-firm trade, trade within the value chain） and finally conclude with our findings to properly account for those trends.
基金grant IDEI-PN-II-ID-PCE-2011-30100 from the Romanian National Science Foundation(ANCS-CNCS)
文摘The basement of the Romanian Carpathians is made of Neoproterozoic to early Paleozoic peri- Gondwanan terranes variably involved in the Variscan orogeny, similarly to other basement terrains of Europe. They were hardly dismembered during the Alpine orogeny and traditionally have their own names in the three Carpathian areas. The Danubian domain of the South Carpathians comprises the Dragsan and Lainici-Paiius peri-Amazonian terranes. The Dragsan terrane originated within the ocean surrounding Rodinia and docked with Rodinia at -800 Ma. It does not contain Cadomian magmatism and consequently it is classified as an Avalonian extra-Cadomian terrane, The Lainici-Pfiius terrane is a Ganderian fragment strongly modified by Cadomian subduction-related magmatism, It is attached to the Moesia platform. The TisoviD terrane is an ophiolite that marks the boundary between Drfagsan and Lainici-Paius terranes. The other basement terranes of the Romanian Carpathians originated close to the Ordovician North- African orogen, as a result of the eastern Rheic Ocean opening and closure. Except for the Sebes-Lotru terrane that includes a lower metamorphic unit of Cadomian age, all the other terranes （Bretila, Tulghes, Negrisoara and Rebra in the East Carpathians, Somes, Biharia and Baia de Aries in the Apuseni mountains, Fagaras, Leaota, Carat and Pades in the South Carpathians） represent late Cambrian-Ordovician rock assemblages. Their provenance, is probably within paleo-nortbeast Africa, close to the Arabian-Nubian shield. The late Cambrian-Ordovician terranes are defined here as Carpathian-type terranes. According to their lithostratigraphy and origin, some are of continental margin magmatic arc setting, whereas others formed in rift and back-arc environment and closed to passive continental margin settings. In a paleo- geographic reconstruction, the continental margin magmatic arc terranes were first that drifted out, followed by the passive continental margin terranes with the back-arc terranes in their front. They accreted to Laurussia during the Variscan orogeny. Some of them （Sebes-Lotru in South Carpathians and Baia de Aries in Apuseni mountains） underwent eclogite-grade metamorphism. The Danubian terranes, the Bretila terrane and the Somes terrane were intruded by Variscan granitoids.
文摘This paper explores the asymmetric effect of COVID-19 pandemic news,as measured by the coronavirus indices(Panic,Hype,Fake News,Sentiment,Infodemic,and Media Coverage),on the cryptocurrency market.Using daily data from January 2020 to September 2021 and the exponential generalized autoregressive conditional heter-oskedasticity model,the results revealed that both adverse and optimistic news had the same effect on Bitcoin returns,indicating fear of missing out behavior does not prevail.Furthermore,when the nonlinear autoregressive distributed lag model is esti-mated,both positive and negative shocks in pandemic indices promote Bitcoin’s daily changes;thus,Bitcoin is resistant to the SARS-CoV-2 pandemic crisis and may serve as a hedge during market turmoil.The analysis of frequency domain causality supports a unidirectional causality running from the Coronavirus Fake News Index and Sentiment Index to Bitcoin returns,whereas daily fluctuations in the Bitcoin price Granger affect the Coronavirus Panic Index and the Hype Index.These findings may have significant policy implications for investors and governments because they highlight the impor-tance of news during turbulent times.The empirical results indicate that pandemic news could significantly influence Bitcoin’s price.
文摘The objective of this work is the study of social and economic inequality in the space of Central and Eastern Europe and its impact on economic growth. Our study includes a three-stage methodology:(1) application of a clustering method based on neural network (Self Organising Maps), to the series of panel data in order to divide countries into clusters, corresponding to the degree of economic and social inequality;(2) computing a composed index of economic and social inequality, using Principal Component Analysis and an extension of the method provided by OECD for computing composite indicators;(3) constructing an econometric model to establish the impact of social and economic inequality on economic growth and a VAR model to determine the causality between main determinants to growth and inequality as well as the response to shocks to the dynamics of the variables. The 24 Eastern and Central European countries have been grouped in five clusters, according to 11 attributes. In the results obtained, the third cluster comprises countries with the most equitable income distribution: Czech Republic, Croatia, Hungary, Slovak Republic, Slovenia. To the opposite side is the fifth cluster with the deepest inequality, including only one country, namely Georgia. The second and third steps of our methodology, were applied only for the extreme clusters namely, the clusters with the highest (C5) and lowest (C3) inequality respectively.
基金Supported by a grant of the Romanian National Authority for Scientific Research,CNCS-UEFISCDI(No.PN-II-ID-PCE-2011-3-0972)
文摘We discussed recent studies, within the framework of transport theories, on heavy ion reactions between charge asymmetric systems, from low up to Fermi energies. We concentrated on the analysis of ternary breakup events of dynamical origin occurring in semi-central reactions, where the formation of excited systems in various conditions of shape, excitation energy and spin is observed. At beam energies around 20 A Me V, we showed how this fragmentation mode emerges from the combined action of surface(neck) instabilities and angular momentum effects, leading to the observation of three aligned massive fragments in the exit channel. At Fermi energies, a transition towards a prompt emission of small fragments from the neck region with larger relative velocity with respect to projectile and target remnants is observed. We also focus on isospin sensitive observables, aiming at extracting information on the density dependence of the isovector part of the nuclear effective interaction and of the nuclear symmetry energy.
基金Analiza impactului incertitudinilor actuale asupra mediului economic,ediția 2022,No.750/19.05.2022(en:Analysis of the impact of current uncertainties on the economic environment,2022 edition,No.750/19.05.2022).Recipient:Cosmin-Octavian CEPOI,PhD.
文摘Using data from the Bucharest Stock Exchange,we examine the factors influencing the probability of informed trading(PIN)during February—October 2020,a COVID-19 pandemic period.Based on an unconditional quantile regression approach,we show that PIN exhibit asymmetric dependency with liquidity and trading costs.Furthermore,building a customized database that contains all insider transactions on the Bucharest Stock Exchange,we reveal that these types of orders monotonically increase the infor-mation asymmetry from the 50th to the 90th quantile throughout the PIN distribution.Finally,we bring strong empirical evidence associating the level of information asym-metry to the level of fake news related to the COVID-19 pandemic.This novel result suggests that during episodes when the level of PIN is medium to high(between 15 and 50%),any COVID-19 related news classified as misinformation released during the lockdown period,is discouraging informed traders to place buy or sell orders condi-tioned by their private information.
文摘The present study will examine the legal regime of authorized natural persons,individual enterprises,and family businesses as regulated in Romania.Such an analysis aims at presenting a panoramic view of the legal framework in a representative country in Eastern Europe.Business activities can be carried out individually by individuals as well as large corporations,firms,and companies that have the status of legal persons.The business carried out by individuals,mainly in terms of the quality of an authorized natural person,is particularly important because it contributes to the development of the middle class within society and to giving professional and financial independence to different categories of professionals.The main aspects related to the qualification of specific notions in the field,the administrative steps for acquiring the status of an authorized natural person,individual enterprise and family enterprise,and their legal regime will be found in the article.
文摘The issue of this study continues the analysis of persuasive truth （the ＂truth＂ of beliefs）, a problem debated by us inseveral articles already published, especially in two of them： Doxastic Dialectics （1999） and The Probable and theProblem （2010）. In this new ＂chapter,＂ our intention is to develop more details from the perspective of subjectivity,which has a grounding role in doxastic dialectics. Doxa＇s axiomatic mechanism tries to temper the subjectivedimension of persuasive truth, by submitting the doxastic proofs to the control of the oppositional principle（=antithetic ＂reasoning＂）. Doxastic thinking discovers dialectically its own ratio （=measure）, progressivelyincreasing the relevance of the listening capacity. In the text that follows, the concept of listening is used in a largerthan sensitive sense, being equivalent to condition of receptivity.
文摘The paper aims to analyze the relationship between economic growth and increasing agricultural productivity at macroeconomic level, relationship that determines a decrease of poverty. Data sets are analyzed using econometric methods in order to test the cointegration relationships. The results of the data show that, in Romania’s case, the added value of agriculture has no significant impact on the evolution of the country’s GDP, so also the impact on raising the standard of living is marginal. To increase productivity and profitability of agriculture requires a complex approach: a financial system to support development of the sector by providing specific financial products and services, the reorganization/retechnologization of farms, encouraging young people to invest in agriculture, ownership structure of agricultural land.
文摘Eye tracking can facilitate understanding irrational decision-making in contexts such as financial risk-taking.For this purpose,we develop an experimental framework in which participants trade a risky asset in a simulated bubble market to maximize individual returns while their eye movements are recorded.Returns are sensitive to eye movement dynamics,depending on the presented visual stimuli.Using eye-tracking data,we investigated the effects of arousal,attention,and disengagement on individual payoffs using linear and nonlinear approaches.By estimating a nonlinear model using attention as a threshold variable,our results suggest that arousal positively influences trading returns,but its effect becomes smaller when attention exceeds a certain threshold,whereas disengagement has a higher negative impact on reduced attention levels and becomes almost irrelevant when attention increases.Hence,we provide a neurobehavioral metric as a function of attention that predicts financial gains in boomand-bust scenarios.This study serves as a proof-of-concept for developing future psychometric measures to enhance decision-making.
文摘The research paper European Tax Models is a comparative member states, taking into account the main features in light analysis of the taxation systems in the European Union of the contribution of indirect and direct taxes and social contributions to the achievement of public revenues. Theme presents a topic of great interest, both theoretically and practically, given that how to place taxes has direct repercussions on the economic development of a country, and undoubtedly influence the rules of an economy, particularly in terms of investment, labor market, and social welfare. It was considered necessary in the first part of the paper to address the conceptual elements and present the most important features of tax systems and the principles that underlie them. It was studied from a theoretical perspective and it found the European tax models as follow： Nordic, continental, Anglo-Saxon, Mediterranean, and catching-up. Then, it analyzed each fiscal European model on each member country, starting from its economic indicators, based on Eurostat data. The objective of the research paper was to present a complete picture of the structure and trends of tax level of the member states of the European Union, sorted by European tax models and the impact of taxation on economic growth and social welfare. The research has concluded that, as long as the rules of the European Union, member states are free to choose their own tax system along with their fiscal policy for economic development and having in a view of their geographical, historical, and political situation.
