Hasil untuk "Banking"

Thomas Katraouras, Dimitrios Rafailidis

Large Language Models (LLMs) have become a mainstay for many everyday applications. However, as data evolve their knowledge quickly becomes outdated. Continual learning aims to update LLMs with new information without erasing previously acquired knowledge. Although methods such as full fine-tuning can incorporate new data, they are computationally expensive and prone to catastrophic forgetting, where prior knowledge is overwritten. Memory-augmented approaches address this by equipping LLMs with a memory bank, that is an external memory module which stores information for future use. However, these methods face a critical limitation, in particular, the memory bank constantly grows in the real-world scenario when large-scale data streams arrive. In this paper, we propose MBC, a model that compresses the memory bank through a codebook optimization strategy during online adaptation learning. To ensure stable learning, we also introduce an online resetting mechanism that prevents codebook collapse. In addition, we employ Key-Value Low-Rank Adaptation in the attention layers of the LLM, enabling efficient utilization of the compressed memory representations. Experiments with benchmark question-answering datasets demonstrate that MBC reduces the memory bank size to 0.3% when compared against the most competitive baseline, while maintaining high retention accuracy during online adaptation learning. Our code is publicly available at https://github.com/Thomkat/MBC.

Susanna Green, Andrew Lundgren

We introduce TemplateGeNN, a fast stochastic template bank generation algorithm which uses Graphical Processing Units (GPUs) and a LearningMatch model (Siamese neural network). TemplateGeNN generated a binary black hole template bank (chirp mass varied from $5 M_{\odot} \leq \mathcal{M}_{c} \leq 20M_{\odot}$, symmetric mass ratio varied from $0.1 \leq η\leq 0.24999$, and equal aligned spin varied from $-0.99 \leq χ_{1,2}\leq 0.99$) of 31,640 templates in $\sim 1$ day on a single A100 GPU. To test the sensitivity of this template bank we injected 7746 binary black hole templates into LIGO Gaussian noise. This template bank recovered 98$\%$ of the injections with a fitting factor greater than 0.97. For lower mass regions (black hole mass region between $5 M_{\odot} \leq m_{1, 2} \leq 25 M_{\odot}$), 99$\%$ of 9469 injections were recovered with a fitting factor greater than 0.97. LearningMatch and TemplateGeNN are a machine-learning pipeline that can be used to accelerate template bank generation for future gravitational-wave data analysis.

João B. G. de Brito, Rodrigo Heldt, Cleo S. Silveira et al.

Financial institutions increasingly adopt customer-centric strategies to enhance profitability and build long-term relationships. While Customer Lifetime Value (CLV) is a core metric, its calculations often rely solely on single-entity data, missing insights from customer activities across multiple firms. This study introduces the Potential Customer Lifetime Value (PCLV) framework, leveraging Open Banking (OB) data to estimate customer value comprehensively. We predict retention probability and estimate Potential Contribution Margins (PCM) from competitor data, enabling PCLV calculation. Results show that OB data can be used to estimate PCLV per competitor, indicating a potential upside of 21.06% over the Actual CLV. PCLV offers a strategic tool for managers to strengthen competitiveness by leveraging OB data and boost profitability by driving marketing efforts at the individual customer level to increase the Actual CLV.

Suhas Vittal, Moinuddin Qureshi

This paper studies the impact of DRAM writes on DDR5-based system. To efficiently perform DRAM writes, modern systems buffer write requests and try to complete multiple write operations whenever the DRAM mode is switched from read to write. When the DRAM system is performing writes, it is not available to service read requests, thus increasing read latency and reducing performance. We observe that, given the presence of on-die ECC in DDR5 devices, the time to perform a write operation varies significantly: from 1x (for writes to banks of different bankgroups) to 6x (for writes to banks within the same bankgroup) to 24x (for conflicting requests to the same bank). If we can orchestrate the write stream to favor write requests that incur lower latency, then we can reduce the stall time from DRAM writes and improve performance. However, for current systems, the write stream is dictated by the cache replacement policy, which makes eviction decisions without being aware of the variable latency of DRAM writes. The key insight of our work is to improve performance by modifying the cache replacement policy to increase bank-parallelism of DRAM writes. Our paper proposes {\em BARD (Bank-Aware Replacement Decisions)}, which modifies the cache replacement policy to favor dirty lines that belong to banks without pending writes. We analyze two variants of BARD: BARD-E (Eviction-based), which changes the eviction policy to evict low-cost dirty lines, and BARD-C (Cleansing-Based), which proactively cleans low-cost dirty lines without modifying the eviction decisions. We develop a hybrid policy (BARD-H), which uses a selective combination of both eviction and writeback. Our evaluations across workloads from SPEC2017, LIGRA, STREAM, and Google server traces show that BARD-H improves performance by 4.3\% on average and up-to 8.5\%. BARD requires only 8 bytes of SRAM per LLC slice.

Deb Narayan Barik, Siddhartha P. Chakrabarty

We present a novel approach for the bank's decision problem, incorporating Limited Liability in the objective function. Accordingly, we consider continuous time models, with and without Limited Liability. We compare the solutions of these two models to demonstrate the effect of inclusion of Limited Liability. To solve the problem with the objective function incorporating Limited Liability, we approximate the payoff function to another set of functions for which we have closed-form solutions. Then, we show that the solution with Limited Liability incorporates less risky assets, while simultaneously increasing the resilience of the bank. After that, we use the metric of $Distance~to~Default$, from the KMV Model, to analyze the bank's resiliency, by considering that the interest rate follows the Vasicek model. Finally, we illustrate the results obtained with a numerical example.

Dan MITRA, Ioana-Florina COITA

This paper investigates how people perceive artificial intelligence (AI), using both original survey data and insights from recent academic and policy literature. As AI technologies become increasingly embedded in daily life—from banking and healthcare to education and justice systems—it is important to understand public sentiment in order to better guide ethical integration and the implementation of effective legislative frameworks. We conducted a survey of 60 individuals with diverse backgrounds to assess their familiarity with AI, perceived benefits and risks, and level of comfort with AI making decisions across various domains. While respondents generally expressed openness toward AI in areas such as transportation and finance, some concerns have emerged around its application in legal proceedings and hiring practices. Some of the key issues included algorithmic bias, erosion of human oversight, and threats to privacy. These results highlight the importance of transparency, public education, and robust governance to correctly align AI deployment with societal expectations and ethical standards.

Md Asad Iqbal Chowdhury, Md Harun Ur Rashid, Mohammad Shamsu Uddin

We offer new insights into how bank efficiency influences the transmission of monetary policy through the bank lending channel (BLC) in an emerging economy context. The study draws on a balanced panel of 29 banks listed on the Dhaka Stock Exchange (DSE), covering the period from 2006 to 2024. We first measure bank efficiency using the non-parametric Data Envelopment Analysis (DEA), which captures both technical efficiency (TE) and scale efficiency (SE). Subsequently, we employ three regression techniques such as Ordinary Least Squares (OLS), Fixed Effects (FE), and the two-step Generalized Method of Moments (GMM) to examine the moderating effects of TE and SE on the relationship between monetary policy instruments—namely, the Money Market Rate (MMR), Treasury Bill Rate (TBR), and Repurchase Agreement Rate (REPO)—and bank lending behaviour. Our findings reveal that banks with higher TE amplify the positive impact of MMR and mitigate the negative effects of TBR, as they effectively manage resources and absorb funding cost shocks. In contrast, banks with higher SE reduce the influence of MMR and exacerbate the contractionary effect of TBR, focusing on cost minimization and margin preservation. The REPO channel shows no significant moderating effect, remaining weak across the models. These findings provide novel evidence that bank efficiency plays a pivotal role in moderating the monetary policy transmission. The study offers important implications for central banks, financial regulators, and policymakers seeking to strengthen the efficacy of monetary transmission by aligning macroeconomic tools with micro-level bank performance dynamics.

Fabian Sven Karst, Sook-Yee Chong, Abigail A. Antenor et al.

The banking sector faces challenges in using deep learning due to data sensitivity and regulatory constraints, but generative AI may offer a solution. Thus, this study identifies effective algorithms for generating synthetic financial transaction data and evaluates five leading models - Conditional Tabular Generative Adversarial Networks (CTGAN), DoppelGANger (DGAN), Wasserstein GAN, Financial Diffusion (FinDiff), and Tabular Variational AutoEncoders (TVAE) - across five criteria: fidelity, synthesis quality, efficiency, privacy, and graph structure. While none of the algorithms is able to replicate the real data's graph structure, each excels in specific areas: DGAN is ideal for privacy-sensitive tasks, FinDiff and TVAE excel in data replication and augmentation, and CTGAN achieves a balance across all five criteria, making it suitable for general applications with moderate privacy concerns. As a result, our findings offer valuable insights for choosing the most suitable algorithm.

Gavin Ruan, Ziqi Guo, Guang Lin

Over 44 million Americans currently suffer from food insecurity, of whom 13 million are children. Across the United States, thousands of food banks and pantries serve as vital sources of food and other forms of aid for food insecure families. By optimizing food bank and pantry locations, food would become more accessible to families who desperately require it. In this work, we introduce a novel two-level optimization framework, which utilizes the K-Medoids clustering algorithm in conjunction with the Open-Source Routing Machine engine, to optimize food bank and pantry locations based on real road distances to houses and house blocks. Our proposed framework also has the adaptability to factor in considerations such as median household income using a pseudo-weighted K-Medoids algorithm. Testing conducted with California and Indiana household data, as well as comparisons with real food bank and pantry locations showed that interestingly, our proposed framework yields food pantry locations superior to those of real existing ones and saves significant distance for households, while there is a marginal penalty on the first level food bank to food pantry distance. Overall, we believe that the second-level benefits of this framework far outweigh any drawbacks and yield a net benefit result.

Islam Abdeljawad, Mamunur Rashid, Muiz Abu Alia et al.

Conventional banks are 'indirectly' allowed to take more risk under the shadow of sovereign guarantees. Banks commit moral hazards as any major banking crisis will be 'cushioned' by deposit insurance and bailed out using the taxpayer's money. This study offers an alternative explanation for the determinants of banks' credit risk, particularly those from the Islamic regions. Although conventional banks and Islamic banks may share state and social cushioning systems, Islamic banks are strictly prohibited by moral and religious principles from gambling with depositors' funds, even if there is a cushion available to bail them out. However, banks belonging to collective societies, such as those in the MENA area, may be inclined to take more risks due to the perception of having a larger safety net to protect them in the event of failure. We analyse these theoretical intersections by utilising a dataset consisting of 320 banks from 20 countries, covering the time span from 2006 to 2021. Our analysis employs a combination of Ordinary Least Squares (OLS), Fixed Effects (FE), and 2-step System-GMM methodologies. Our analysis reveals that Islamic banks are less exposed to credit risk compared to conventional banks. We contend that the stricter ethical and moral ground and multi-layer monitoring system amid protracted geopolitical and post-pandemic crises impacting Islamic countries contribute to the lower credit risk. We examine the consequences for credit and liquidity management in Islamic banks and the risk management strategies employed by Islamic banks, which can serve as a valuable reference for other banks.

Ahmad Nurkhin, Fachrurrozie, Anna Kania Widiatami et al.

This paper examines how good corporate governance (GCG) affects Indonesian banks’ non-performing loans (NPLs) and its relevance to the current banking sector situation in Indonesia. The research findings provide a comprehensive understanding of the effect of bank-specific factors on NPLs, offering timely and important insights for the banking industry. This quantitative study focuses on commercial banks listed on the Indonesian Stock Exchange in 2021. The observation period spans four years (2018–2021), utilizing 216-unit panel data from 54 banks for analysis. Documentation was used for data collection, and panel data multiple regression analysis was employed as the data analysis technique. The findings indicate that increased board of directors’ meetings are associated with higher NPLs, while having independent board commissioners correlates with lower NPLs. The p-value of the board of director meetings is 0.027, and the coefficient is 0.005037. The p-value of the board of independent board commissioners is 0.017, and the coefficient is –0.00109. Effective GCG implementation is crucial in maintaining credit quality and reducing NPL levels. The p-value of the GCG score is 0.043, and the coefficient is –0.42985. However, the frequency of Board of Commissioners’ meetings does not significantly affect NPLs. The study also shows that the Loan Deposit Ratio (LDR) and bank size negatively and significantly impact NPLs. In contrast, Return on Equity (ROE) and leverage do not significantly affect NPL levels in Indonesian banks. This study provides empirical evidence that underscores the importance of robust GCG, especially during the challenging business conditions triggered by the pandemic. AcknowledgmentsThis study received funding from LPPM UNNES, contract number 12.12.4/UN37/PPK.10/2023.

Jürg M. Blum

Yanjie Sun, Zhike Gong, Quan Shi et al.

Effective control of credit risk is a key link in the steady operation of commercial banks. This paper is mainly based on the customer information dataset of a foreign commercial bank in Kaggle, and we use LightGBM algorithm to build a classifier to classify customers, to help the bank judge the possibility of customer credit default. This paper mainly deals with characteristic engineering, such as missing value processing, coding, imbalanced samples, etc., which greatly improves the machine learning effect. The main innovation of this paper is to construct new feature attributes on the basis of the original dataset so that the accuracy of the classifier reaches 0.734, and the AUC reaches 0.772, which is more than many classifiers based on the same dataset. The model can provide some reference for commercial banks' credit granting, and also provide some feature processing ideas for other similar studies.

José Alejandro Fernández Fernández

AbstractThis article begins with an analysis of banking flows in the euro zone, through a complex network, from 2006 to 2020. This analysis allows us to observe the topology of the network through different phases of the business cycle. It is obtained that there is greater fragmentation in the network that increases in three stages, with turning points in crises. In turn, the topological structure is less random and presents more capitalized subnetworks with less risk. As for the nodes of the network, Germany gives up the position of centrality in favor of France. The determinants of the links in the network are analyzed with Machine Learning, obtaining as push and pull bank variables solvency and bank income structure, respectively, and productivity as economic variable.

Umbaruk Kieta Phanum

Banking as an extensive information subject is constantly changing under the marketing influences of the era of huge information. Checking out the revolutionary important specifics analytic sources, as Data Mining techniques are essential for the banking business. This aims to reveal valuable information from the vast length of info and realize far better strategic management in addition to client pleasure. To provide decent assistance for future research and development, a most thorough present analysis of the existing study problem of DM in banking will be unbelievably beneficial. Since pre existing ratings just tackle the applications until 2013, this specific newspaper seeks to run this specific exploration gap, and also supplies the sizable progressions and most current DM implementations in banking document 2013. By collecting and analyzing the fads of review concentration, info online resources, specialized aids, and info analytical online resources, this specific newspaper helps obtain important insights about the succeeding developments of equally DM together with the banking business, in addition to a comprehensive one stop guide table. Furthermore, we recognize the main obstacles and also provide a summary for all those interested in big data.

Nadiia Davydenko, Yuliya Lutsyk, Alina Buriak et al.

The article is devoted to the modern development of high technologies and computer technology greatly enhanced the development of automated banking systems of banking sector organizations and allowed the synthesis of information and communication technologies for their formation.The main purpose of the article is to select the main indicators for assessing the level of financial security of the banking system of the state and identify promising areas of its development using forecasting models. In the process of research such analytical functions have been used: polynomial, exponential, power and logarithmic. The authors believe that the information and analytical provision of the financial security of the bank is an information provision that combines, on the one hand, information work, that is, ways, means and methods of collecting the necessary information, and on the other - analytical work, which includes forms and methods of information analysis and processing, which ensures an objective assessment of the situation and the adoption of a balanced management decision. As a result, forecast models were built for each of the indicators and also, it has been found that most indicators of the banking system of Ukraine in 2021-2023 will remain at “unsatisfactory” and “critical” levels. In conclusions it was proposed to introduce measures that would be aimed at improving the reliability and stability of the banking system of Ukraine.

Svitlana Khalatur, Svitlana Kachula, Vitalii Oleksiuk et al.

Crisis management is an important tool for managing modern agricultural businesses, especially in the face of uncertainty and changes in the market. This article examines the role of crisis management as a key element in the formation of a financial mechanism for the sustainable development of the agricultural sector. It analyses the main aspects of crisis management in agricultural business and its impact on the formation of a sustainable financial mechanism. The relationship between crisis management and sustainable development of the agrarian sector is studied. The possibilities of using the principles of crisis management to improve the financial stability and competitiveness of agricultural enterprises are determined. As a result, the article emphasizes the importance of crisis management as a key factor in the formation of a sustainable financial mechanism for achieving sustainable development of agricultural businesses. As follows, the scientific novelty in the article lies in several key aspects: integration of crisis management and sustainable development; application of crisis management principles to agriculture; emphasis on financial mechanisms: the article focuses on the financial aspect of crisis management and sustainable development in agriculture. Thus, the scientific novelty of the article lies in its innovative approach to integrating crisis management principles into the agricultural context, emphasizing the financial mechanism involved in the pursuit of sustainable development in the agricultural sector. The results of the study can be useful for agricultural entrepreneurs, managers, academics, and regulators to improve management strategies and increase the sustainability of the agricultural sector.

Linda Allen, A. Rai

L. Schover, K. Brey, A. Lichtin et al.

C. Matutes, X. Vives