Abstract The Basel Accord emphasizes the necessity of employing internal data models to manage key credit risk components, including Probability of Default (PD), Loss Given Default (LGD), and Exposure At Default (EAD). Among these, internal datasets are critical for estimating PD, a fundamental measure of borrower creditworthiness. Nevertheless, practical application often faces challenges due to incomplete datasets, which can skew analyses and undermine the accuracy of credit scoring models. Traditional approaches to addressing missing data, such as sample deletion or mean imputation, are widely used; however, they often prove insufficient for accurate prediction. Consequently, imputation methods are typically favored over deletion, as they allow for the full utilization of available data. Recent advancements have introduced more sophisticated techniques, such as Generative Adversarial Imputation Networks (GAIN), which utilize a generative adversarial network to model data distributions and impute missing values with greater precision than conventional methods. Building on these developments, this study proposes a novel imputation approach, SMART (Structured Missingness Analysis and Reconstruction Technique) for credit scoring datasets. SMART consists of two primary stages: first, it normalizes and denoises the dataset using randomized Singular Value Decomposition (rSVD), followed by the implementation of GAIN to impute missing values. Experimental results demonstrate that SMART significantly outperforms existing state-of-the-art methods, particularly in high missing data contexts (20%, 50%, and 80%), with improvements in imputation accuracy of 7.04%, 6.34%, and 13.38%, respectively. In conclusion, SMART represents a substantial advancement in handling incomplete credit scoring datasets, leading to more precise PD estimation and enhancing the robustness of credit risk management models.
Managing issue reports is essential for the evolution and maintenance of software systems. However, manual issue management tasks such as triaging, prioritizing, localizing, and resolving issues are highly resource-intensive for projects with large codebases and users. To address this challenge, we present SPRINT, a GitHub application that utilizes state-of-the-art deep learning techniques to streamline issue management tasks. SPRINT assists developers by: (i) identifying existing issues similar to newly reported ones, (ii) predicting issue severity, and (iii) suggesting code files that likely require modification to solve the issues. We evaluated SPRINT using existing datasets and methodologies, measuring its predictive performance, and conducted a user study with five professional developers to assess its usability and usefulness. The results show that SPRINT is accurate, usable, and useful, providing evidence of its effectiveness in assisting developers in managing issue reports. SPRINT is an open-source tool available at https://github.com/sea-lab-wm/sprint_issue_report_assistant_tool.
As organizations increasingly rely on data-driven insights, the ability to run data intensive applications seamlessly across multiple cloud environments becomes critical for tapping into cloud innovations while complying with various security and regulatory requirements. However, big data application development and deployment remain challenging to accomplish in such environments. With the increasing containerization and modernization of big data applications, we argue that a unified control/management plane now makes sense for running these applications in hybrid cloud environments. To this end, we study the problem of building a generic hybrid-cloud management plane to radically simplify managing big data applications. A generic architecture for hybrid-cloud management, called Titchener, is proposed in this paper. Titchener comprises of independent and loosely coupled local control planes interacting with a highly available public cloud hosted global management plane. We describe a possible instantiation of Titchener based on Kubernetes and address issues related to global service discovery, network connectivity and access control enforcement. We also validate our proposed designs with a real management plane implementation based on a popular big data workflow orchestration in hybrid-cloud environments.
Abubeker Abdurahman, Abrar Hossain, Kevin A Brown
et al.
Efficient job scheduling and resource management contribute towards system throughput and efficiency maximization in high-performance computing (HPC) systems. In this paper, we introduce a scalable job scheduling and resource management component within the structural simulation toolkit (SST), a cycle-accurate and parallel discrete-event simulator. Our proposed simulator includes state-of-the-art job scheduling algorithms and resource management techniques. Additionally, it introduces workflow management components that support the simulation of task dependencies and resource allocations, crucial for workflows typical in scientific computing and data-intensive applications. We present the validation and scalability results of our job scheduling simulator. Simulation shows that our simulator achieves good accuracy in various metrics (e.g., job wait times, number of nodes usage) and also achieves good parallel performance.
IntroductionChaotic resonance is similar to stochastic resonance, which emerges from chaos as an internal dynamical fluctuation. In chaotic resonance, chaos-chaos intermittency (CCI), in which the chaotic orbits shift between the separated attractor regions, synchronizes with a weak input signal. Chaotic resonance exhibits higher sensitivity than stochastic resonance. However, engineering applications are difficult because adjusting the internal system parameters, especially of biological systems, to induce chaotic resonance from the outside environment is challenging. Moreover, several studies reported abnormal neural activity caused by CCI. Recently, our study proposed that the double-Gaussian-filtered reduced region of orbit (RRO) method (abbreviated as DG-RRO), using external feedback signals to generate chaotic resonance, could control CCI with a lower perturbation strength than the conventional RRO method.MethodThis study applied the DG-RRO method to a model which includes excitatory and inhibitory neuron populations in the frontal cortex as typical neural systems with CCI behavior.Results and discussionOur results reveal that DG-RRO can be applied to neural systems with extremely low perturbation but still maintain robust effectiveness compared to conventional RRO, even in noisy environments.
This massive development of information technology makes it easier for people's lives in various fields, one of them is social media, social media that people use a lot to get information about news or events that are happening in Indonesia, one of which is social media Twitter which provides a lot of information for the people of Indonesia, one of which is information about Covid-19 which is currently rife in the territory of Indonesia Sentiment analysis is a branch of Natural Language Processing (NLP) which can help determine the sentiments that occur in society. This study uses data in the form of tweets to carry out sentiment analysis obtained on Twitter social media.This research utilizes one of the Supervised Learning algorithms, namely Support Vector Machine. In this study, three (3) kernels are used for the Support Vector Machine, each of which is Linear, Radial basis function and Polynomial, to find which kernel produces the highest accuracy value. From the experiments carried out using data sharing for training as much as 70% and for testing data as much as 30% of the total data of 6000 data, the resulting accuracy value for the Support Vector Machine method on the Linear kernel produces an accuracy value of 89% and for the Radial kernel base function accuracy by 90% and for the Polynomial kernel it produces an accuracy of 88%. So it is concluded for the three (3) kernels for testing the Support Vector Machine method on the Radial basis function kernel to produce the best accuracy value
Recently, 6G in-X subnetworks have been proposed as low-power short-range radio cells to support localized extreme wireless connectivity inside entities such as industrial robots, vehicles, and the human body. Deployment of in-X subnetworks within these entities may result in rapid changes in interference levels and thus, varying link quality. This paper investigates distributed dynamic channel allocation to mitigate inter-subnetwork interference in dense in-factory deployments of 6G in-X subnetworks. This paper introduces two new techniques, Federated Multi-Agent Double Deep Q-Network (F-MADDQN) and Federated Multi-Agent Deep Proximal Policy Optimization (F-MADPPO), for channel allocation in 6G in-X subnetworks. These techniques are based on a client-to-server horizontal federated reinforcement learning framework. The methods require sharing only local model weights with a centralized gNB for federated aggregation thereby preserving local data privacy and security. Simulations were conducted using a practical indoor factory environment proposed by 5G-ACIA and 3GPP models for in-factory environments. The results showed that the proposed methods achieved slightly better performance than baseline schemes with significantly reduced signaling overhead compared to the baseline solutions. The schemes also showed better robustness and generalization ability to changes in deployment densities and propagation parameters.
Crop production management is essential for optimizing yield and minimizing a field's environmental impact to crop fields, yet it remains challenging due to the complex and stochastic processes involved. Recently, researchers have turned to machine learning to address these complexities. Specifically, reinforcement learning (RL), a cutting-edge approach designed to learn optimal decision-making strategies through trial and error in dynamic environments, has emerged as a promising tool for developing adaptive crop management policies. RL models aim to optimize long-term rewards by continuously interacting with the environment, making them well-suited for tackling the uncertainties and variability inherent in crop management. Studies have shown that RL can generate crop management policies that compete with, and even outperform, expert-designed policies within simulation-based crop models. In the gym-DSSAT crop model environment, one of the most widely used simulators for crop management, proximal policy optimization (PPO) and deep Q-networks (DQN) have shown promising results. However, these methods have not yet been systematically evaluated under identical conditions. In this study, we evaluated PPO and DQN against static baseline policies across three different RL tasks, fertilization, irrigation, and mixed management, provided by the gym-DSSAT environment. To ensure a fair comparison, we used consistent default parameters, identical reward functions, and the same environment settings. Our results indicate that PPO outperforms DQN in fertilization and irrigation tasks, while DQN excels in the mixed management task. This comparative analysis provides critical insights into the strengths and limitations of each approach, advancing the development of more effective RL-based crop management strategies.
Trigeneration provides an effective means of power, heat, and cold production on site. Proper design and well-managed operation of such units can bring in substantial savings in consumed primary energy as well as in the amount of greenhouse gases released to the atmosphere, compared to separate production of all three media. The studied sub-MW-sized trigeneration unit comprises an internal combustion engine combined with an absorption chiller and a heat management system, delivering all three media to a nearby industrial facility. A mathematical model is developed based on available design and process data, a profit function is set up, and the subsequent sensitivity analysis of economic parameters is realized. The lowered efficiency of summer operation is analyzed, and a suitable solution is proposed, with an estimated total investment cost of EUR 114,000 and an anticipated simple payback period less than 2 years.
Hale Pamukçu, Pelin Soyertaş Yapıcıoğlu, Mehmet İrfan Yeşilnacar
This study majorly aimed to determine the effect of optimization on transport routes on the reduction of greenhouse gas (GHG) emissions from municipal solid waste management (MSM) within the scope of European Union (EU) Green Deal. Optimization of collection and transportation routes has been regarded as an effective methodology in order to mitigate the GHG emissions of municipal waste management, recently. Optimization of routes has been obtained using ant colony algorithm (ACA) and Monte Carlo simulation, in this study. In this context, this study investigated to reduce GHG emissions from municipal waste management using optimization of transportation routes in Diyarbakir city in Turkey. Firstly, GHG emissions which are carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O) emissions from waste collection and transport have been calculated using a new developed model based on Tier-I method. Then, Monte Carlo simulation has been used to figure out the GHG emissions. Finally, life cycle assessment (LCA) approach has been applied to determine the GHG emissions. According to the route optimization resulting ACA methodology, nearly 47.43% of reduction on each GHG emissions. Approximately, 58%, 38% and 51% of reduction on CO2, CH4 and N2O emissions respectively has been achieved, in the result of the route optimization using Monte Carlo simulation. The results of LCA methodology revealed that the reduction reached up 45.71% on GHG emissions in terms of Global Warming Potential (GWP). The reduction amounts have been overlapped with each other.
Ilaria Finore, Luigi Leone, Alessia Gioiello
et al.
Abstract Background The management of the organic waste recycling process determines the interest in the thermophiles microorganisms involved in composting. Although many microbial enzymes have been isolated and studied for their industrial and commercial uses, there is still a continuous search for microorganisms which could synthesize industrially feasible enzymes, especially when the microbial diversity of cow dung itself makes a potential source of biotechnological enzymes. Results The composting process studied at the Experimental Station of the University of Naples Federico II (Castel Volturno, Caserta, Italy) was characterized by fresh saw dust 40%, bovine manure 58%, and 2% mature compost as raw organic substrates, and its thermophilic phase exceeded a temperature of 55 °C for at least 5 days, thus achieving sanitation. Six microbial strains were isolated and designated as follow: CV1-1, CV1-2, CV2-1, CV2-2, CV2-3 and CV2-4. Based on 16S rRNA gene sequence, HRMAS–NMR spectroscopy, and biochemical investigations, they were ascribed to the genera Geobacillus and Bacillus. All the microbial isolates were qualitatively screened on plates for the presence of hydrolytic activities, and they were quantitatively screened in liquid for glycoside hydrolase enzymes in the extracellular, cell-bound, and cytosolic fractions. Based on these results, strains CV2-1 and CV2-3 were also quantitatively screened for the presence of cellulase and pectinase activities, and pH and temperature optimum plus thermostability of cellulase from CV2-1 were analyzed. Conclusions The isolation and the identification of these thermophilic microorganisms such as Geobacillus toebii, Geobacillus galactosidasius, Bacillus composti, Bacillus thermophilus and Aeribacillus composti have allowed the study of the biodiversity of compost, with emphasis on their primary metabolome through an innovative and underutilized technique, that is HRMAS–NMR, also highlighting it as a novel approach to bacterial cell analysis. Subsequently, this study has permitted the identification of enzymatic activities able to degrade cellulose and other polymeric substrates, such as the one investigated from strain CV2-1, which could be interesting from an industrial and a biotechnological point of view, furthermore, increasing the knowledge for potential applicability in different industrial fields as an efficient and environmentally friendly technique. Graphical Abstract
Introduction/Main Objectives: The study proposed is written based on the results of quantitative research and the analysis of the theory and practice of leadership. The study's main objective is to determine the essential traits of a leader for effective interaction with team members. Background Problems: Most research on this topic chose a leader's traits based on analyzing literary sources rather than on empirical research. Novelty: The traits for the most effective collaboration between leader and team members were chosen by potential and actual members of the leader's team, namely students and teachers of the University. Research Methods: We conducted a questionnaire survey of 103 teachers and 421 Bogomolets National Medical University (Kyiv) students. The statistical analysis was carried out using Wald Test. Finding/Results: The research confirmed that both respondent categories admitted the importance of all leadership traits. At the same time, such traits as passion, effectiveness, self-confidence, determination, and ability to take risks appeared to be more significant for the students than for the teachers. The teachers ranked such a trait as decency higher than the students did. Also, such issues as the importance of organizational culture, ethical aspects of leadership, and the most effective leadership style for productive interaction with team members were examined. Conclusion: This study proposed complex recommendations for creating the most productive model of the interaction between the leader and team members based on the data obtained.
Bastian Greisner, Dieter Mauer, Frank Rögener
et al.
This study investigated the predictability of forward osmosis (FO) performance with an unknown feed solution composition, which is important in industrial applications where process solutions are concentrated but their composition is unknown. A fit function of the unknown solution’s osmotic pressure was created, correlating it with the recovery rate, limited by solubility. The osmotic concentration was derived and used in the subsequent simulation of the permeate flux in the considered FO membrane. For comparison, magnesium chloride and magnesium sulfate solutions were used since these show a particularly strong deviation from the ideal osmotic pressure according to Van’t Hoff and are, thus, characterized by an osmotic coefficient unequal to 1. The simulation is based on the solution–diffusion model with consideration of external and internal concentration polarization phenomena. Here, a membrane module was subdivided into 25 segments of equal membrane area, and the module performance was solved by a numerical differential. Experiments in a laboratory scale for validation confirmed that the simulation gave satisfactory results. The recovery rate in the experimental run could be described for both solutions with a relative error of less than 5%, while the calculated water flux as a mathematical derivative of the recovery rate showed a bigger deviation.
Hyder Akmal S., Sundström Agneta, Chowdhury Ehsanul H.
Purpose: This study explores the knowledge development of network-based market orientation (MO) for the internationalization of disruptive innovation (DI) by small and medium-sized enterprises (SMEs).
Knowledge management systems (KMS) are in high demand for industrial researchers, chemical or research enterprises, or evidence-based decision making. However, existing systems have limitations in categorizing and organizing paper insights or relationships. Traditional databases are usually disjoint with logging systems, which limit its utility in generating concise, collated overviews. In this work, we briefly survey existing approaches of this problem space and propose a unified framework that utilizes relational databases to log hierarchical information to facilitate the research and writing process, or generate useful knowledge from references or insights from connected concepts. Our framework of bidirectional knowledge management system (BKMS) enables novel functionalities encompassing improved hierarchical note-taking, AI-assisted brainstorming, and multi-directional relationships. Potential applications include managing inventories and changes for manufacture or research enterprises, or generating analytic reports with evidence-based decision making.
This paper focuses on efficient landmark management in radar based simultaneous localization and mapping (SLAM). Landmark management is necessary in order to maintain a consistent map of the estimated landmarks relative to the estimate of the platform's pose. This task is particularly important when faced with multiple detections from the same landmark and/or dynamic environments where the location of a landmark can change. A further challenge with radar data is the presence of false detections. Accordingly, we propose a simple yet efficient rule based solution for radar SLAM landmark management. Assuming a low-dynamic environment, there are several steps in our solution: new landmarks need to be detected and included, false landmarks need to be identified and removed, and the consistency of the landmarks registered in the map needs to be maintained. To illustrate our solution, we run an extended Kalman filter SLAM algorithm in an environment containing both stationary and temporally stationary landmarks. Our simulation results demonstrate that the proposed solution is capable of reliably managing landmarks even when faced with false detections and multiple detections from the same landmark.
Joanna Katarzyna Banach, Ryszard Żywica, Paulius Matusevičius
Among the challenges of sustainable management of meat production, the key issue is to improve the energy efficiency of production processes, which will consequently affect the reduction of greenhouse gas emissions. Such effects are achieved by combining various chilling systems with electrical stimulation that determines the quality of meat at the slaughter stage. The novelties of the research undertaken included determining the impact of various variants of meat production (chilling method: slow, fast, accelerated + HVES/NES) on changes in the basic (industrial) quality indicators (pH and temperature) of beef produced from Polish Holstein-Friesian breed cattle, and then indicating the optimal variant for energy-efficient (sustainable) beef production. The HVES and the fast chilling method yielded positive economic (meat weight loss), technological (high quality, hot-boning), energetic (lower electricity consumption), and organizational effects (reduced chilling and storage surfaces and expenditures for staff wages) compared to the slow and accelerated methods. Reaching the desired final temperature with an increased amount of chilled meat enables obtaining a few-fold decrease in the specific energy consumption and a higher energy efficiency of the process. This allows recommending the above actions to be undertaken by entrepreneurs in the pursuit of sustainable meat production.