Resource Usage Research Articles

Multisegment Overlap–Save Method for Coherent Dedispersion

Dispersion occurs due to the interstellar medium, which functions as a prism and causes different time delays in radio waves of varying frequencies. The coherent dedispersion technique is often used in pulsar and fast radio burst observations to mitigate this phenomenon. The widely recognized Overlap–Save approach enables this dedispersion algorithm to efficiently perform long linear convolution. However, with the present implementation, the necessary filter length for dedispersion can reach 100 million points or more. The corresponding fast Fourier transform (FFT) points should be larger than this value and a GPU cluster can be used to tackle this demanding process. This study presents the Multisegment Overlap–Save Method (MS-OSM) to effectively address this problem. Our algorithm divides signal bands into separate short segments based on frequency component delays during dedispersion. By using the short segments shuffling technique with the Overlap-Save structure, MS-OSM can greatly reduce the FFT and inverse FFT points required down to fewer than 65,536 points. To evaluate the performance of MS-OSM, a synthetic pulsar signal is created and verified using standard software tools like DSPSR and PRESTO. The results show that MS-OSM maintains the same resolution while reducing execution time and resource usage. The validation of MS-OSM is taken by processing real pulsar observation data.

Reducing Makespan and Enhancing Resource Usage in Cloud Computing With ESJFP Method: A New Dynamic Approach

ABSTRACTA well‐designed web‐based tool or application allows consumers to access on‐demand services on a pay‐per‐use basis via the internet in the cloud computing service paradigm. Cloud computing continues to be a leading technology trend, with a primary focus on optimizing and enhancing end‐user applications. The increasing challenge of meeting the diverse needs of clints for service providers is propelling the development of load scheduling algorithms. Some of the current scheduling algorithms used in cloud computing include First Come First Serve (FCFS), Shortest Job First (SJF) and Round Robin (RR). The response time and makespan—the interval between the start and end times of consecutive tasks on the same machine—are the two most crucial load balancing elements. Response time refers to the duration a server takes to respond to a client's request. Measured in milliseconds, this timer starts when a client sends a request and stops when the server sends its initial response. This study examines many load balancing methods and suggests improvements for cloud computing's Shortest Job First (SJF) methodology. To minimize makespan and maximize resource usage, we provide a solution that is Enhanced Shortest Job First with Priority (ESJFP) load scheduling algorithms. Under the ESJFP method, computers with more processing power are assigned the longest tasks with higher MIPS (million instructions per second) needs, while machines with lesser processing capacity are assigned the shortest jobs with lower MIPS requirements. By giving equal priority to all activities, this technique makes sure that neither high‐MIPS nor low‐MIPS jobs have to wait an extended period of time for resource allocation.

A Multi-Class ECG Signal Classifier Using a Binarized Depthwise Separable CNN with the Merged Convolution–Pooling Method

Binarized convolutional neural networks (bCNNs) are favored for the design of low-storage, low-power cardiac arrhythmia classifiers owing to their high weight compression rate. However, multi-class classification of ECG signals based on bCNNs is challenging due to the accuracy loss introduced by the binarization operation. In this paper, an effective multi-classifier system is proposed for electrocardiogram (ECG) signals using a binarized depthwise separable convolutional neural network (bDSCNN) with the merged convolution–pooling (MCP) method. The binarized depthwise separable convolution layer is adopted to reduce the increased number of parameters in multi-classification systems. Instead of operating convolution and pooling sequentially as in a traditional convolutional neural network (CNN), the MCP method merges pooling together with convolution layers to reduce the number of computations. To further reduce hardware resources, this work employs blockwise incremental calculation to eliminate redundant storage with computations. In addition, the R peak interval data are integrated with P-QRS-T features to improve the classification accuracy. The proposed bDSCNN model is evaluated on an Intel DE1-SoC field-programmable gate array (FPGA), and the experimental results demonstrate that the proposed system achieves a five-class classification accuracy of 96.61% and a macro-F1 score of 89.08%, along with a dynamic power dissipation of 20 μW for five-category ECG signal classification. The hardware resource usage of BRAM and LUTs plus REGs is reduced by at least 2.94 and 1.74 times, respectively, compared with existing ECG classifiers using bCNN methods.

User perspectives on library digitization and its impact on research capabilities

This study assessed user perspectives on library digitization and its impact on research capabilities. A survey design with both quantitative and qualitative approaches was employed. Simple random sampling was used to select respondents whereas purposive sampling was used to select key informants. Data were collected using online questionnaires and interviews. MS Excel was used to analyze quantitative data while qualitative data was analyzed using thematic method. Findings show that the majority of the respondents utilize library resources frequently with statistical analysis (MS = 0.387998, t = −1.45939, p < 0.146074) indicating there is no significant difference between education level and the frequency of library resources usage. Also, the majority of respondents have positive perceptions towards the use of library resources with t-statistic (6.599663291, is above 1.649982976 for one-tail and 1.967956506 for two-tail) indicating a significant difference between age and perceptions towards the use of library resources. Moreover, findings show capacity of a single document to be used by many users at once and information to be accessed at a fingerprint are major impacts of library digitization. Furthermore, findings reveal issues related to licensing, subscription costs, and copyright and overwhelming volume of information available as the major challenges user face when utilizing digital library resources. This study marks the pioneering exploration into the impacts of library digitization on research capabilities. The study therefore, recommends a continuous training to students to be able to access and critically evaluate the overwhelming amount of online information. Also, users should be trained on how to navigate copyright issues when accessing digital library resources. Additionally, libraries should consider expanding their collections by incorporating more open access resources. Moreover, parent institutions should make sure that they increase the budget to their digital libraries so that libraries can increase the subscriptions of all useful resources.

Experimental assessment of containers running on top of virtual machines

AbstractOver the past two decades, the cloud computing paradigm has gradually attracted more popularity due to its efficient resource usage and simple service access model. Virtualisation technology is the fundamental element of cloud computing that brings several benefits to cloud users and providers, such as workload isolation, energy efficiency, server consolidation, and cost reduction. This paper examines the combination of operating system‐level virtualisation (containers) and hardware‐level virtualisation (virtual machines). To this end, the performance of containers running on top of virtual machines is experimentally compared with standalone virtual machines and containers based on different hardware resources, including the processor, main memory, disk, and network in a real testbed by running the most commonly used benchmarks. Paravirtualisation and full virtualisation as well as type 1 and type 2 hypervisors are covered in this study. In addition, three prevalent containerisation platforms are examined.

GTBTL-IoT: An Approach of Curtailing Task Offloading Time for Improved Responsiveness in IoT-MEC Model

INTRODUCTION: The Internet of Things (IoT) has transformed daily life by interconnecting digital devices via integrated sensors, software, and connectivity. Although IoT devices excel at real-time data collection and decision-making, their performance on complex tasks is hindered by limited power, resources, and time. To address this, IoT is often combined with cloud computing (CC) to meet time-sensitive demands. However, the distance between IoT devices and cloud servers can result in latency issues. OBJECTIVES: To mitigate latency challenges, Mobile Edge Computing (MEC) is integrated with IoT. MEC offers cloud-like services through servers located near network edges and IoT devices, enhancing device responsiveness by reducing transmission and processing latency. This study aims to develop a solution to optimize task offloading in IoT-MEC environments, addressing challenges like latency, uneven workloads, and network congestion. METHODS: This research introduces the Game Theory-Based Task Latency (GTBTL-IoT) algorithm, a two-way task offloading approach employing Game Matching Theory and Data Partitioning Theory. Initially, the algorithm matches IoT devices with the nearest MEC server using game-matching theory. Subsequently, it splits the entire task into two halves and allocates them to both local and MEC servers for parallel computation, optimizing resource usage and workload balance. RESULTS: GTBTL-IoT outperforms existing algorithms, such as the Delay-Aware Online Workload Allocation (DAOWA) Algorithm, Fuzzy Algorithm (FA), and Dynamic Task Scheduling (DTS), by an average of 143.75 ms with a 5.5 s system deadline. Additionally, it significantly reduces task transmission, computation latency, and overall job offloading time by 59%. Evaluated in an ENIGMA-based simulation environment, GTBTL-IoT demonstrates its ability to compute requests in real-time with optimal resource usage, ensuring efficient and balanced task execution in the IoT-MEC paradigm. CONCLUSION: The Game Theory-Based Task Latency (GTBTL-IoT) algorithm presents a novel approach to optimize task offloading in IoT-MEC environments. By leveraging Game Matching Theory and Data Partitioning Theory, GTBTL-IoT effectively reduces latency, balances workloads, and optimizes resource usage. The algorithm's superior performance compared to existing methods underscores its potential to enhance the responsiveness and efficiency of IoT devices in real-world applications, ensuring seamless task execution in IoT-MEC systems.

Method for Rail Surface Defect Detection Based on Neural Network Architecture Search

Abstract This study addresses the inherent limitations of implementing neural network architecture search algorithms for rail surface defect detection, including low search efficiency and the oversight of edge features on the rail surface. A sophisticated multi-level neural network architecture search framework is proposed that integrates and emphasizes rail surface edge features. The framework utilizes the Z-Score normalization method to quantify the edge concern of rail surface defect samples, combined with an Edge-Loss function to enhance edge feature recognition capabilities. Furthermore, acknowledging the sensitivity of defect features to spatial resolution changes, a multi-level neural network architecture search space is meticulously designed. In the cell-level search space, a method combining partial channel sampling with operation pruning is employed to enhance model search efficiency and regularization. In the network-level search space, optimal paths for resolution change are established, allowing for the screening and aggregation of defect features at various levels to facilitate the adaptive extraction of multi-scale edge defect features. Experimental outcomes indicate that this method significantly reduces computational resource usage by approximately 75% and increases mIOU by 2.6% relative to traditional architecture search methods. Moreover, it demonstrates robust capability in accurately recognizing defective edges on rail surfaces, thereby substantiating the method's effectiveness.

EFFICIENCY OF URBAN PASSENGER TRANSPORT BY BUS: AN APPLICATION TO THE CASE OF CUENCA-ECUADOR

Objective: The study aims to analyze the technical and scale efficiency of public transportation in Cuenca, Ecuador, using Data Envelopment Analysis (DEA). It evaluates both supply and demand aspects. Theoretical Framework: DEA measures the efficiency of decision-making units (DMUs) in using inputs to produce outputs. It's widely used to assess public transportation efficiency by comparing resource usage and service outcomes. Method: The study analyzes 37 DMUs, considering kilometers traveled, number of frequencies, and total fleet as inputs, and operating costs and passenger demand as outputs. Results and Discussion: DEA_Demand showed a 90% technical efficiency and 45% scale efficiency. DEA_Offer showed 94% and 96% efficiency, respectively, highlighting the potential for adjustments in both approaches. Research Implications: The findings help policymakers optimize resource allocation and improve public transport services. Originality/Value: The study uniquely combines supply and demand analysis using DEA in a mid-sized city like Cuenca, offering actionable insights.

Automated decision making in Barrett’s oesophagus: development and deployment of a natural language processing tool

Manual decisions regarding the timing of surveillance endoscopy for premalignant Barrett’s oesophagus (BO) is error-prone. This leads to inefficient resource usage and safety risks. To automate decision-making, we fine-tuned Bidirectional Encoder Representations from Transformers (BERT) models to categorize BO length (EndoBERT) and worst histopathological grade (PathBERT) on 4,831 endoscopy and 4,581 pathology reports from Guy’s and St Thomas’ Hospital (GSTT). The accuracies for EndoBERT test sets from GSTT, King’s College Hospital (KCH), and Sandwell and West Birmingham Hospitals (SWB) were 0.95, 0.86, and 0.99, respectively. Average accuracies for PathBERT were 0.93, 0.91, and 0.92, respectively. A retrospective analysis of 1640 GSTT reports revealed a 27% discrepancy between endoscopists’ decisions and model recommendations. This study underscores the development and deployment of NLP-based software in BO surveillance, demonstrating high performance at multiple sites. The analysis emphasizes the potential efficiency of automation in enhancing precision and guideline adherence in clinical decision-making.

Reducing the cost of cold start time in serverless function executions using granularity trees

In serverless computing, cold starts significantly impede performance. This paper presents a granularity tree-based scheduling strategy, dynamically adjusting serverless function deployment by package dependencies to mitigate cold starts and optimize resource usage. This approach notably reduces cold start and response times. Empirical results from evaluating functions across various datasets show the strategy outperforms existing methods. Specifically, it consistently delivers lower response times and decreases resource consumption, demonstrating its effectiveness in managing computational resources while ensuring swift function invocation. In particular scenarios, the proposed scheduler impressively reduced response times from 8134.1 ms to 392.8 ms and idle memory usage from 15.2 GB to 11.2 GB per machine. In other scenarios, it reduced response times from 12,152.7 ms to 504.2 ms while maintaining a 100% function execution percentage. These quantified improvements underscore the significant enhancements in cold start mitigation and overall system performance, highlighting the potential of granularity tree-based scheduling in enhancing serverless computing architectures by effectively balancing rapid response with reduced resource usage.

Prediction of the impact of tobacco waste hydrothermal products on compost microbial growth using hyperspectral imaging combined with machine learning

The insufficient understanding of the impact of hydrothermal products on the growth characteristics of compost microorganisms presents a significant challenge to the broader implementation of hydrothermal coupled composting for tobacco waste. Traditional biochemical detection methods are labor-intensive and time-consuming, highlighting the need for faster and more accurate alternatives. This study investigated the effects of hydrothermal treatment on tobacco straw products and their influence on compost microorganism growth, using hyperspectral imaging (HSI) technology and machine learning algorithms. Sixty-one tobacco straw samples were analyzed with a hyperspectral camera, and image processing was used to extract average spectra from regions of interest (ROI). Hierarchical cluster analysis (HCA) and principal component analysis (PCA) were applied to assess four key variables: nicotine content, total humic acid content, Penicillium chrysogenum H/C ratio, and Bacillus subtilis OD600 ratio. The effects of hydrothermal treatment on compost were classified as promoting, inhibiting, or neutral regarding microbial growth. The Competitive Adaptive Reweighted Sampling (CARS) method identified the most influential wavelengths in the 900-1700 nm spectral range. The Random Forest (RF) model outperformed SVM, KNN, and XGBoost models in predicting microbial growth responses, achieving Rc = 0.957, RMSE = 3.584. Key wavelengths were identified at 1096 nm, 1101 nm, 1163 nm, 1335 nm, and 1421 nm. The results indicate that hyperspectral imaging combined with machine learning can accurately predict changes in the chemical composition of tobacco straws and their effects on microbial activity. This method provides an innovative and effective means of improving the resource usage of tobacco straws in composting, enhancing sustainable waste management procedures.

Modelling and optimization of experimental parameters for porosity in friction stir processed aluminum foam using response surface methodology

ABSTRACT Aluminium foams are an emerging multifunctional material with a distinct structure and unique properties. However, controlling the porosity and pore morphology is still challenging because of the dependency on a number of the process parameters. This study uses a statistical approach to control porosity and influence of each parameter on porosity in a friction-stir processed aluminium plate. A second-order Response Surface Methodology (RSM) model is developed, and relationship among porosity and process parameters like Weight % TiH2, Weight % Al2O3, number of passes, tool rpm, and foaming temperature has been established and influence of each parameter on porosity is investigated. Experimental results validate the model, optimising 80–85% porosity for multifunctional applications. The optimal parameters are 1.2 wt% TiH2, 2.3 wt% Al2O3, 3 passes, 3000 tool rpm, and 738 °C foaming temperature. It was observed from the ANOVA that most influencing parameter is tool rpm, number of passes, followed by foaming temperature. The Scanning electron microscopy (SEM) and Energy-dispersive X-ray spectroscopy (EDS) analyses of the fabricated foam at its optimised condition confirm the uniform distribution of pores and the elements. This statistical optimisation model minimises experimental time and resource usage, demonstrating substantial technical and industrial significance.

Observer‐based PID control for fuzzy dynamic systems with memory triggering protocol

AbstractThis paper addresses the issue of observer‐based proportional–integral–derivative control for Takagi–Sugeno fuzzy dynamic systems using a memory triggering protocol. To tackle the challenges of managing nonlinear systems, the Takagi–Sugeno fuzzy approach is employed to approximate these systems with a series of local linear models. A novel memory event‐triggering mechanism is introduced, utilizing a computer's storage unit to store historical data and adjust the triggering threshold based on this data, thereby reducing unnecessary network resource usage. Given that the system state can be affected by external disturbances and become unpredictable, an observer‐based proportional–integral–derivative controller is designed to manage these uncertainties. Finally, a car‐damper‐spring model is presented to validate the proposed methodology.

Robust deadline-aware network function parallelization framework under demand uncertainty

The orchestration of Service Function Chains (SFCs) in Mobile Edge Computing (MEC) becomes crucial for ensuring efficient service provision, especially under dynamic and uncertain demand. Meanwhile, the parallelization of Virtual Network Functions (VNFs) within an SFC can further optimize resource usage and reduce the risk of deadline violations. However, most existing works formulate the SFC orchestration problem in MEC with deterministic demands and costly runtime resource reprovisioning to handle dynamic demands. This paper introduces a Robust Deadline-aware network function Parallelization framework under Demand Uncertainty (RDPDU) designed to address the challenges posed by unpredictable fluctuations in user demand and resource availability within MEC networks. RDPDU to consider end-to-end latency for SFC assembly by modeling load-dependent processing latency and load-independent propagation latency. Also, RDPDU formulates the problem assuming uncertain demand by Quadratic Integer Programming (QIP) to be resistant to dynamic service demand fluctuations. By discovering dependencies between VNFs, the RDPDU effectively assembles multiple sub-SFCs instead of the original SFC. Finally, our framework uses Deep Reinforcement Learning (DRL) to assemble sub-SFCs with guaranteed latency and deadline. By integrating DRL into the SFC orchestration problem, the framework adapts to changing network conditions and demand patterns, improving the overall system's flexibility and robustness. Experimental evaluations show that the proposed framework can effectively deal with demand fluctuations, latency, deadline, and scalability and improve performance against recent algorithms.

UAVs Revolutionizing Efficiency: Integrating DRL and Random Walrus for Enhanced Energy and Spectral Resource Management

The improvement of energy efficiency and spectral efficiency in networks is achieved by seamlessly integrating energy harvesting, cognitive radio technologies, and Non-Orthogonal Multiple Access (NOMA) techniques. These complementary strategies optimize resource usage and address challenges related to energy consumption. Additionally, the adaptability and versatility of Unmanned Aerial Vehicles (UAVs) offer innovative solutions for enhancing coverage performance, thereby improving connectivity, efficiency, and reliability. We introduce a novel approach called the Deep Reinforcement Learning-Random Walrus (DRL-RW) algorithm, which combines Deep Reinforcement Learning (DRL) with the Random Walrus optimization (RWO) technique for efficient spectrum resource allocation and energy harvesting management in dynamic environments. The DRL-RW algorithm enables UAVs to learn optimal spectrum-sharing strategies and energy harvesting policies, while the RWO enhances the algorithm's adaptability and speed in exploring diverse solutions. Simulation results demonstrate the effectiveness of the DRL-RW algorithm, showing significant improvements in several performance metrics, including reduced energy consumption, enhanced computation time, improved convergence, increased signal-to-noise ratio, higher throughput, extended network lifetime, and increased harvested energy. These findings underscore the efficacy of the DRL-RW approach in addressing challenges associated with energy management in cognitive radio networks. The integration of UAVs, NOMA networks, and this novel algorithm represents a promising direction for developing energy-efficient communication systems.

Multiobjective Hybrid Al-Biruni Earth Namib Beetle Optimization and Deep Learning based Task Scheduling in Cloud Computing

With the rapid development of computing networks, cloud computing (CC) enables the deployment of large-scale applications and meets the increased rate of computational demands. Moreover, task scheduling is an essential process in CC. The tasks must be effectually scheduled across the Virtual Machines (VMs) to increase resource usage and diminish the makespan. In this paper, the multi-objective optimization called Al-Biruni Earth Namib Beetle Optimization (BENBO) with the Bidirectional-Long Short-Term Memory (Bi-LSTM) named as BENBO+ Bi-LSTM is developed for Task scheduling. The user task is subjected to the multi-objective BENBO, in which parameters like makespan, computational cost, reliability, and predicted energy are used to schedule the task. Simultaneously, the user task is fed to Bi-LSTM-based task scheduling, in which the VM parameters like average computation cost, Earliest Starting Time (EST), task priority, and Earliest Finishing Time (EFT) as well as the task parameters like bandwidth and memory capacity are utilized to schedule the task. Moreover, the task scheduling outcomes from the multi-objective BENBO and Bi-LSTM are fused for obtaining the final scheduling with less makespan and resource usage. Moreover, the predicted energy, resource utilization and makespan are considered to validate the BENBO+ Bi-LSTM-based task scheduling, which offered the optimal values of 0.669J, 0.535 and 0.381.

Comparative analysis of performance, efficiency, and resource usage between COMSOL Multiphysics and the MPh library of Python in the simulation of physical phenomena

MPh is a library of Python that makes possible to link the Python computer language with COMSOL Multiphysics. The use of the MPh library opens the possibility to save the computer resources employed when simulating physical phenomena and solving mathematical models and equations. In the Python command interpreter is possible to change or adjust some settings and parameters from the models created in COMSOL, and to execute the COMSOL kernel to solve those models. In this study, we compare the performance of COMSOL Multiphysics and the MPh library of Python when computing the magnetic field generated by a distribution of currents and ferromagnetic material. The metrics employed to do the comparison and the methodology to measure them are described, as well as the computer equipment where the programs ran. The results show that the execution time of the computations are similar in both software, but in the rest of the metrics, the execution in Python surpassed the execution in COMSOL. We showed that the MPh library of Python demands less CPU and RAM usage when solving the mathematical models and the files containing the solutions use less storage memory. As a conclusion, we propose the use of the MPh library of Python to improve the performance of the computer in charge of carrying out the calculations.

Analysis on the Influence law and Mechanism of the Strength of Reclaimed Aggregate-Industrial Waste Slag Solidified Silt

To optimize the usage of construction waste resources, industrial waste slag and silt, this paper used Portland cement, mineral waste residue and phosphogypsum composite to make cementing material (CMPS) with construction waste reclaimed aggregate to solidify silt. The laboratory solidification test and microscopic analysis were conducted to assess the mechanical characteristics of solidified sludge. In order to clarify the mineral composition, microscopic morphology and pore characteristics of the regenerated aggregate and CMPS solidified silt (SS), XRD, SEM and nitrogen adsorption pore analyzer were employed for further explore the regenerated aggregate and CMPS solidified silt effectively, and further reveal the internal mechanism of the regenerated aggregate and CMPS solidified silt effectively. The findings indicated that the strength of Portland cementmineral waste residue phosphogypsum terpolymer system curing agent increased by 107.34% than that of single Port-land cement SS at 56 d, and the strength of CMPS solidified silt increased by 25.68% under the action of reclaimed aggregate framework. The maintenance period and water composition of the silt have high correlation with the strength. Therefore, influence law of above two influencing factors on its mechanical properties were further explored and the strength prediction were made. The microscopic test results showed that, drawing from the hydration of Portland cement and pozzolans reaction of mineral waste residue, the solidified system has pre-pared CaSiO3 hydrate gel and ettringite crystals with gelatinous properties, which help to fill the pores and form a denser structure.

Automatic selection of the best performing control point approach for project control with resource constraints

During project execution, the actual project progress shows deviations from the baseline schedule due to uncertainty. To complete the project timely, project monitoring is performed at discrete control points to identify project opportunities/problems and take possible corrective actions. These control points affect the quality of project monitoring and corrective actions, but little guidance is available on identifying situations where the control points pay off the most in terms of project duration.This paper proposes new control point approaches considering the risk, the complexity of the network, and subnetwork information to determine the timing of project monitoring and action taking. Moreover, new parameters are proposed to model more realistic project characteristics. Subsequently, a classification model is developed to select the best performing control point approach given project characteristics. An extensive computational experiment is conducted on a set of 3,810 artificial projects with diverse project characteristics to evaluate the performance of the classification model and further validate it on empirical project data.The computational results indicate that the classification model outperforms the average performance of any proposed control point approaches. The results also show that the resource variability that indicates the resource usage deviations between project activities is the primary driver for detecting the best control point approach for projects with resource constraints.

Optimizing multi-time series forecasting for enhanced cloud resource utilization based on machine learning

Due to its flexibility, cloud computing has become essential in modern operational schemes. However, the effective management of cloud resources to ensure cost-effectiveness and maintain high performance presents significant challenges. The pay-as-you-go pricing model, while convenient, can lead to escalated expenses and hinder long-term planning. Consequently, FinOps advocates proactive management strategies, with resource usage prediction emerging as a crucial optimization category. In this research, we introduce the multi-time series forecasting system (MSFS), a novel approach for data-driven resource optimization alongside the hybrid ensemble anomaly detection algorithm (HEADA). Our method prioritizes the concept-centric approach, focusing on factors such as prediction uncertainty, interpretability and domain-specific measures. Furthermore, we introduce the similarity-based time-series grouping (STG) method as a core component of MSFS for optimizing multi-time series forecasting, ensuring its scalability with the rapid growth of the cloud environment. The experiments performed demonstrate that our group-specific forecasting model (GSFM) approach enabled MSFS to achieve a significant cost reduction of up to 44%.