Decision-making Strategies Research Articles

Enhancing breast cancer prediction through stacking ensemble and deep learning integration

Breast cancer is one of the most common types of cancer in women and is recognized as a serious global public health issue. The increasing incidence of breast cancer emphasizes the importance of early detection, which enhances the effectiveness of treatment processes. In addressing this challenge, the importance of machine learning and deep learning technologies is increasingly recognized. The aim of this study is to evaluate the integration of ensemble models and deep learning models using stacking ensemble techniques on the Breast Cancer Wisconsin (Diagnostic) dataset and to enhance breast cancer diagnosis through this methodology. To achieve this, the efficacy of ensemble methods such as Random Forest, XGBoost, LightGBM, ExtraTrees, HistGradientBoosting, AdaBoost, GradientBoosting, and CatBoost in modeling breast cancer diagnosis was comprehensively evaluated. In addition to ensemble methods, deep learning models including convolutional neural network (CNN), recurrent neural network (RNN), gated recurrent unit (GRU), bidirectional long short-term memory (BILSTM), long short-term memory (LSTM) were analyzed as meta predictors. Among these models, CNN stood out for its high accuracy and rapid training time, making it an ideal choice for real-time diagnostic applications. Finally, the study demonstrated how breast cancer prediction was enhanced by integrating a set of base predictors, such as LightGBM, ExtraTrees, and CatBoost, with a deep learning-based meta-predictor, such as CNN, using stacking ensemble methodology. This stacking integration model offers significant potential for healthcare decision support systems with high accuracy, F1 score, and receiver operating characteristic area under the curve (ROC AUC), along with reduced training times. The results from this research offer important insights for enhancing decision-making strategies in the diagnosis and management of breast cancer.

Navigating Treatment Ambiguities in Head and Neck Squamous Cell Carcinoma: A Retrospective Analysis of Equivocal PET/CT Findings and Decision-Making Strategies Following Primary Non-Surgical Treatment

Navigating Treatment Ambiguities in Head and Neck Squamous Cell Carcinoma: A Retrospective Analysis of Equivocal PET/CT Findings and Decision-Making Strategies Following Primary Non-Surgical Treatment

The Feasibility Of Diversification Strategies On The Transformation Of Coal Mining Companies Into Mineral Mining Companies In Indonesia

Following the global commitment on net zero emissions during the Paris Agreement in 2016, Indonesia is fulfilling its commitment by focusing their strategy on electric vehicles batteries investment and decreasing the coal-based power generation. One of the company groups in Indonesia that mainly operating on coal starts to look at green minerals that would increase in demand as a diversification strategy. This study aims to investigate which green mineral options that would be the most feasible strategy for the company through decision-making method. Based on the literature review on business strategy and decision-making method decision tree analysis, a questionnaire was distributed across the internal team of the target company that works on the diversification attempt. The respondents were asked on their confidence on the proposed green mineral options. Analysis on the responses shows that there is a strong, if not perfect, relationship between the confidence scores and all the calculated expected monetary values (EMV) of the proposed green minerals. The analysis also shows that the highest EMV compared to other green mineral option, is gold. On this basis, it is recommended for the company to choose gold as their diversification strategy on green minerals. Further research is needed to investigate other factors that has not been covered in this dissertation, or to investigate through other decision-making method.

Geohazard Plugin: A QGIS Plugin for the Preliminary Analysis of Landslides at Medium–Small Scale

Landslides are a major global threat, endangering lives, infrastructure, and economies. This paper introduces the Geohazard plugin, an open-source tool for QGIS, designed to support medium–small-scale landslide analysis and management. The plugin integrates several algorithms, including the Groundmotion–C index for evaluating SAR data reliability, Landslide–Shalstab for assessing shallow landslide susceptibility, and Rockfall–Droka for estimating rockfall invasion areas and the rockfall relative (spatial) hazard. An application example is provided for each module to facilitate validation and discussion. A case study from the Western Italian Alps highlights the practical application of the Rockfall–Droka modules, showcasing their potential to identify critical zones by integrating the results on affected areas, process intensity, and preferential paths. Emphasis is given to the calibration of model parameters, a critical aspect of the analysis, achieved through a back-analysis of a rockfall event that occurred in June 2024. The Geohazard plugin streamlines geohazard assessments, providing land managers with actionable insights for decision-making and risk mitigation strategies. This user-friendly GIS tool contributes to enhancing resilience in landslide-prone regions.

Green Suppliers Performance Analysis in A Supply Chain Netwotk using Fuzzy Topsis and Fuzzy Vikor Techniques

Green Supply Chain Management (GSCM) strives to decrease, if not eliminate, supply chain operations' negative environmental effects.Multi-criteria decision-making (MCDM) strategies should be utilized to evaluate suppliers' GSCM performance because GSCM includes multi-dimensional methods. In order to supply solutions to perplexing and demanding multi-attribute inquiries in a fuzzy environment, fuzzy group decision-making procedures must be produced. Based on integrated fuzzy MCDM methodologies, this research provides a methodology for assessing the GSCM performance of firms in terms of green design, green image, green transformation, green logistics, and green management system. The fuzzy DEMATEL technique is used to calculate the cause and effect relationships between GSCM dimensions. Based on this association, the fuzzy AHP approach is utilized to generate the weights of the relevant criterion.

Effectuation, causation and CEO tenure: navigating SME performance in dynamic environments

Purpose Amidst global economic uncertainties, small and medium-sized enterprises (SMEs) are confronted with the pressing need to adapt to complex and ambiguous external landscapes. This study aims to investigate how the decision-making strategies of causation and effectuation are related to the performance of Russian SMEs and whether and how this relationship is moderated by CEO tenure. Design/methodology/approach Using a sample of 602 Russian SMEs, hierarchical linear modeling and its estimation using ordinary least squares is conducted. Robustness checks and post hoc analysis are performed. Findings The findings provide insights into effectuation theory by looking at and testing its basic tenets in an underexplored context and investigating the direct and moderated effects of causation and effectuation on SME performance. Moreover, the study builds upon upper-echelons theory (UET) by considering CEO tenure – an individual-level characteristic – as a moderator in the effectuation/causation and SME performance relationship. A framework for understanding how executives’ attributes shape organizational strategies and outcomes is provided. Originality/value The findings contribute to effectuation theory by testing its core principles in emerging market context, examining both direct and moderated effects of causation and effectuation on SME performance. Moreover, the study builds upon upper-echelons theory (UET) by considering CEO tenure – an individual-level characteristic – as a moderator in the effectuation/causation and SME performance relationship. This study offers a framework for understanding how executive characteristics shape organizational strategies and outcomes in emerging markets.

Beyond Visual Appeal: The Impact of Multisensory Experience of Hotel Marketing and Review Images on Sales

Images, with their direct content experience, have become essential in tourists’ narratives of accommodation sharing. Although visual content plays an important role in influencing user decisions, research on hotel images often neglects the significance of sensory elements. Sensory elements in visual content help shape tourists’ perceptions and evaluations of their accommodation experiences. This study explores how sensory elements in marketing and review images affect hotel sales from a sensory perspective. The research employs a combination of hybrid machine vision and econometric methods to examine how visual, tactile, auditory, and smell elements influence hotel sales. These results reveal the relationship among sensory elements in marketing and review images, their discrepancies, and hotel sales. The primary theoretical significance of this study is in analyzing how sensory elements in image content influence hotel sales, providing a unique sensory perspective for tourist decision-making and hotel marketing strategies.

Tacit knowledge-informed approximate dynamic programming for last-mile delivery operations in online-to-offline pharmacies

PurposeTacit knowledge in frontline operations is primarily reflected in the holders’ intuition about dynamic systems. Despite the implicit nature of tacit knowledge, the understanding of complex systems it encapsulates can be displayed through formalization methods. This study seeks to develop a methodology for formalizing tacit knowledge in a dynamic delivery system.Design/methodology/approachThis study employs a structured survey to gather experiential knowledge from dispatchers engaged in last-mile delivery operations. This knowledge is then formalized using a value function approximation approach, which transforms tacit insights into structured inputs for dynamic decision-making. We apply this methodology to optimize delivery operations in an online-to-offline pharmacy context.FindingsThe raw system feature data are not strongly correlated with the system’s development trends, making them ineffective for guiding dynamic decision-making. However, the system features obtained through preprocessing the raw data increase the predictiveness of dynamic decisions and improve the overall effectiveness of decision-making in delivery operations.Research limitations/implicationsThis research provides a foundational framework for studying sequential dynamic decision problems, highlighting the potential for improved decision quality and system optimization through the formalization and integration of tacit knowledge.Practical implicationsThis approach proposed in this study offers a method to preserve and formalize critical operational expertise. By embedding tacit knowledge into decision-making systems, organizations can enhance real-time responsiveness and reduce operational costs.Originality/valueThis study presents a novel approach to integrating tacit knowledge into dynamic decision-making frameworks, demonstrated in a real-world last-mile delivery context. Unlike previous research that focuses primarily on explicit data-driven methods, our approach leverages the implicit, experience-based insights of operational staff, leading to more informed and effective decision-making strategies.

Individual differences in training time in the rat gambling task are unrelated to subsequent decision-making strategies

IntroductionDecision-making requires individuals to perceive probabilities and risks associated with different options. The Iowa gambling task (IGT) is a widely used instrument that assesses decision-making under uncertainty and risk by varying monetary reinforcer/loss contingencies. The rat gambling task (rGT), based on the IGT, is a preclinical test using varying number of palatable reinforcers as wins and different duration of timeouts as punishment, mimicking losses. The rGT requires extensive operant training prior to the free choice sessions. The aim of the present study was to investigate if task acquisition and number of training days affected subsequent individual differences in decision-making strategies in the rGT, and if behavioral profiles impacted on task learning.MethodTraining time and performance of 70 male Lister Hooded rats from previously published studies were herein used to investigate whether learning time affected later decision-making strategies in the free choice rGT. Behavioral profiles generated from a subset of animals were used to study the impact of underlying behavior on learning time.ResultsThere were differences in training days between fast, intermediate and slow learners. However, time required to acquire the rGT did not affect subsequent decision-making strategies in the free choice rGT. Finally, learning time was independent of underlying behavioral profiles.DiscussionIn conclusion, neither decision-making strategies in the rGT nor behavioral profiles were correlated or differed between animals with different learning speed. This suggests that the large variation in training time between animals is unrelated to subsequent decision-making strategies during free choice rGT. Such information is valuable for researchers using the rGT.

Impact of Negative Decisions on Mental Health among College and University Students

The purpose of the study is to explore the intricate relationship between negative decisions and the mental health of college and university students. The correlation research design will be used for this study. The sample of the study will comprise N=300, evenly distributed between men (n=150) and women (n=150) in a university setting that will be selected ensuring the inclusion and exclusion criteria. Participants were selected through random sampling. The decision-making questionnaire (DMQ) was developed by Paul T. Coster (1998) and consists of 21 items, General health questionnaire (GHQ) was developed by David Goldberg (1994), and consisted of items 12. The finding aims to uncover the impact of negative experiences affects same as males and females. It is hypothesized that there will be a significant relationship between negative decision-making and mental health. Or there is a significant positive relationship between bad experiences or mental health among university or college students. The findings have significant implications for institutions to prioritize the students’ decision-making strategies or create a supportive environment to enhance their well-being to test these hypotheses Pearson moment correlation or t-test will measure Data will analyzed using the SPSS 20.0 version.

Insurance-Service-Oriented Maintenance Strategy of Rural Highway Slopes under Typhoon Disasters

Rural highway slopes are characterized by poor resistance to natural disasters and low insurance coverage, which poses a significant challenge for insurance companies. This paper aims to construct an insurance-oriented rural highway slope maintenance decision-making framework to optimize the slope maintenance strategy and effectively reduce the insurance company’s claim costs. Based on the expectation theory, a set of flexible calculation methods for slope maintenance cost and post-disaster reconstruction cost is proposed. Combined with the probability of slope disaster occurrence, the corresponding maintenance decision-making scheme is formulated. Through the analysis of practical engineering cases in China, the effectiveness of the proposed method is verified. The results show that, after calculating the maintenance strategy of 184 slopes, it is found that nine of the slopes are maintained before the typhoon disaster, and it is expected that claims costs can be saved of up to 44.5 million Chinese yuan. This study not only provides a new way to improve the income of insurance companies, but also provides theoretical support and practical guidance for disaster risk management of rural highway slopes.

Machine learning insights into early mortality risks for small cell lung cancer patients post-chemotherapy

IntroductionSmall cell lung cancer (SCLC) is a highly aggressive form of lung cancer, and chemotherapy remains a cornerstone of its management. However, the treatment is associated with significant risks, including heightened toxicity and early mortality. This study aimed to quantify the 90-day mortality rate post-chemotherapy in SCLC patients, identify associated features, and develop a predictive machine learning model.MethodsThis study utilized data from the Surveillance, Epidemiology, and End Results (SEER) database (2000–2018) to identify prognostic features influencing early mortality in SCLC patients. Prognostic features were selected through univariate logistic regression and Lasso analyses. Predictive modeling was performed using advanced machine learning algorithms, including XGBoost, Multilayer Perceptron, K-Nearest Neighbor, and Random Forest. Additionally, traditional models, such as logistic regression and AJCC staging, were employed for comparison. Model performance was evaluated using key metrics, including the Area Under the Receiver Operating Characteristic Curve (AUC), calibration plots, the Kolmogorov–Smirnov (KS) statistic, and Decision Curve Analysis (DCA).ResultsAnalysis of 12,500 eligible patients revealed 10 clinical features significantly impacting outcomes. The XGBoost model demonstrated superior discriminatory capability, achieving AUC scores of 0.95 in the training set and 0.78 in the validation set. It outperformed comparative models across all datasets, as evidenced by its AUC, KS score, calibration, and DCA results. Additionally, the model was integrated into a web-based platform to improve accessibility.ConclusionThis study introduces a machine learning model alongside a web-based support system as critical resources for healthcare professionals, facilitating personalized clinical decision-making and enhancing treatment strategies for SCLC patients post-chemotherapy.

Surgeon attitudes toward risk stratification in emergency surgery for the elderly: an ESTES cross-sectional survey.

Our study explores the utilization of objective tools for preoperative assessment of elderly patients by Emergency General Surgeons (EGS). A descriptive cross-sectional survey was conducted via the European Society for Trauma and Emergency Surgery (ESTES) Research Committee. EGS were invited through the ESTES members' mailing list and social media platforms. The survey included two sections: (1) clinical scenarios involving elderly patients with varying chronic conditions, and (2) participant characteristics. Data collection lasted 12 weeks, with reminders sent every 4 weeks. Statistical analyses were performed using Microsoft Excel and EasyMedStat. One hundred and seven surgeons responded to the survey. Median respondent age was 41 years, with a male prevalence (72.9%). Most participants were from Europe (85%). Key-findings included that 62.6% reported using one or more risk assessment tools (RATs), while 35.5% used one or more frailty scores. Additionally, 4.7% were unaware of any RATs, and 35.5% were unaware of any frailty scores. Decision-making strategies leveraging personal experience with minimal impact from RATs predominated. Preoperative risk assessment tool and frailty score use for elderly patients requiring emergency surgery remains limited among ESTES surgeons. Our study highlights the need for focused education and tool workflow integration to improve risk stratification, decision-making and outcomes. Institutional approaches coupled with targeted educational interventions using implementation science principles are recommended to bridge this knowledge-to-action gap. Future research should focus on developing comprehensive, user-friendly tools and evaluating their impact on patient-centered outcomes.

Introduction to behavioral science and its practical applications in public health.

The World Health Organization (WHO) highlights the importance of improving the design of public health interventions and policies by applying principles from the behavioral sciences. These sciences play a crucial role in modifying behaviors and addressing a wide range of health challenges, from pandemics and chronic diseases to the climate crisis. This article examines the transformative impact of behavioral sciences on public health promotion, focusing on the factors that influence decision-making and intervention strategies based on the six principles developed by the WHO Technical Advisory Group. Additionally, it addresses the significant shortage of Spanish-language literature on this topic, reviewing contributions from influential scientists and key theoretical models. The WHO's recommendations for the effective implementation of these interventions in public health are also discussed. This work not only fills a critical gap in the Spanish-language literature but also provides practical tools to improve public health in Spanish-speaking countries.

Development of per Capita GDP Forecasting Model Using Deep Learning: Including Consumer Goods Index and Unemployment Rate

In the 21st century, the increasing complexity and uncertainty of the global economy have heightened the need for accurate economic forecasting. Per capita GDP, a critical indicator of living standards, economic growth, and productivity, plays a key role in government policy-making, corporate strategy, and investor decisions. However, predicting per capita GDP poses significant challenges due to its sensitivity to various economic and social factors. Traditional methods such as statistical analysis, regression, and time-series models have shown limitations in capturing nonlinear interactions and volatility of economic data. To address these limitations, this study develops a per capita GDP forecasting model based on deep learning, incorporating key macroeconomic variables—the Consumer Price Index (CPI) and unemployment rate (UR)—to enhance predictive accuracy. This study employs five deep-learning regression models (RNN, LSTM, GRU, TCN, and Transformer) applied to real and placebo datasets, each incorporating combinations of CPI and UR. The results demonstrate that deep learning models can effectively capture complex, nonlinear relationships in economic data, significantly improving predictive accuracy compared to traditional models. Among the models, the Transformer consistently achieves the highest R-squared and lowest error values across various metrics (MSE, RMSE, and MSLE), indicating its superior ability to model intricate economic patterns. In addition, including CPI and UR as additional predictors enhances model robustness, with the TCN and Transformer models showing particularly strong performance in capturing short-term economic fluctuations. The findings suggest that the deep learning models, especially the Transformer, offer valuable tools for policymakers and business leaders, providing reliable GDP forecasts that support economic decision-making, resource allocation, and strategic planning. Academically, this study advances the understanding of deep learning applications in economic forecasting, particularly in integrating significant macroeconomic variables for enhanced predictive performance. The developed model is a foundation for informed economic policy and strategic decisions, offering a robust and actionable framework for managing economic uncertainties. This research contributes to theoretical and applied economics, providing insights that bridge academic innovation with practical utility in economic forecasting.

Configurations for high outsourcing performance in construction projects: an integrated perspective of transaction costs and capabilities

PurposeThis study aims to integrate perspectives of transaction costs and capabilities to investigate how the configurational fit between outsourcing strategies, asset specificity of projects and the capabilities of general contractors could achieve high outsourcing performance.Design/methodology/approachEmploying fuzzy-set qualitative comparative analysis, this study analyzes 31 outsourcing cases in construction projects to examine outsourcing performance under different combinations of transaction costs and capabilities.FindingsThe findings highlight six different but functionally equivalent configurations to high outsourcing performance, which vary according to attributes of projects and capabilities of contractors. Further, this study develops four context-dependent paths that link outsourcing strategies to high outsourcing performance: leading practitioner, collaborative manager, partnership dependent, and struggler.Practical implicationsHow do contractors in construction projects navigate outsourcing practices to high performance? This study provides a configurational perspective to make outsourcing strategy decisions. Based on the firm’s capabilities and project asset specificity, contractors are encouraged to select from six different but functionally equivalent combinations and thereby four outsourcing strategies, namely leading practitioner, collaborative manager, partnership dependent, and struggler, to achieve high outsourcing performance.Originality/valueThis study contributes to the body of knowledge by offering a holistic perspective that integrates production and transaction costs to analyze contractors’ outsourcing strategies, contributing to the integration of the transaction cost perspective and capability perspective.

Evaluating variability in decision-making among spine surgeons treating adult spine deformity.

Significant variability in the management of Adult Spinal Deformity (ASD) has been observed among spine surgeons worldwide. The variability among Canadian spine surgeons, a country with universal public healthcare, remains unknown. The study aims to evaluate areas of variability in the perioperative optimization and surgical management of ASD among Canadian spine surgeons. In this cross-sectional study, 25 Canadian spine surgeons, predominantly orthopedic surgeons (20) and neurosurgeons (5) with varying experience, participated in an online survey focused on Adult Spinal Deformity (ASD). The study involved 25 Canadian spine surgeons with varying level of experience, representing both orthopedic and neurosurgical specialities. The study aimed to evaluate the variability in surgical decision-making and perioperative optimization strategies among Canadian spine surgeons when faced with simulated scenarios of ASD pathologies. The online survey presented 4 vignettes with simulated scenarios of the most common ASD pathologies, including High Grade Spondylolisthesis (HGS), Neglected Adolescent Idiopathic Scoliosis (NAIS), Degenerative Scoliosis (DS), and Flat Back Syndrome (FBS). Questions in the vignettes explored ASD surgical decision-making, while additional questions focused on perioperative optimization. Descriptive statistics were used to analyze multiple-choice responses, and open-text responses were categorized into themes. Variability was observed in the duration conservative treatment across the 4 ASD cases. Surgeons exhibited variability in the use of preoperative osteoporosis treatment. There was varied use of a dedicated anesthesiology team. Surgical goals varied in HGS and NAIS. The primary surgical method was variable in DS and HGS, the type of osteotomy varied in DS and FBS, and level of fixation varied in HGS and NAIS. Consensus was observed in the use of intraoperative monitoring across of all 4 ASD pathologies, the implementation of a team-based approach, and the selection of the primary surgical goal in DS and FBS. Our cross-sectional study revealed variability among Canadian spine surgeons in the management of ASD, potentially influenced by the uncertain ASD progression, the need for evidence-based nonsurgical guidelines, and insufficient evidence on optimal surgical approaches. These findings will help guide future research to ultimately reduce variability and improve ASD patient management and outcomes.

Digital Age Transformation in Patient-Physician Communication: 25-Year Narrative Review (1999-2023).

The evolution of patient-physician communication has changed since the emergence of the World Wide Web. Health information technology (health IT) has become an influential tool, providing patients with access to a breadth of health information electronically. While such information has greatly facilitated communication between patients and physicians, it has also led to information overload and the potential for spreading misinformation. This could potentially result in suboptimal health care outcomes for patients. In the digital age, effectively integrating health IT with patient empowerment, strong patient-physician relationships, and shared decision-making could be increasingly important for health communication and reduce these risks. This review aims to identify key factors in health communication and demonstrate how essential elements in the communication model, such as health IT, patient empowerment, and shared decision-making, can be utilized to optimize patient-physician communication and, ultimately, improve patient outcomes in the digital age. Databases including PubMed, Web of Science, Scopus, PsycINFO, and IEEE Xplore were searched using keywords related to patient empowerment, health IT, shared decision-making, patient-physician relationship, and health communication for studies published between 1999 and 2023. The data were constrained by a modified query using a multidatabase search strategy. The screening process was supported by the web-based software tool Rayyan. The review methodology involved carefully designed steps to provide a comprehensive summary of existing research. Topic modeling, trend analysis, and synthesis were applied to analyze and evaluate topics, trends, and gaps in health communication. From a total of 389 selected studies, topic modeling analysis identified 3 primary topics: (1) Patient-Physician Relationship and Shared Decision-Making, (2) Patient Empowerment and Education Strategies, and (3) Health Care Systems and Health IT Implementations. Trend analysis further indicated their frequency and prominence in health communication from 1999 to 2023. Detailed examinations were conducted using secondary terms, including trust, health IT, patient-physician relationship, and patient empowerment, derived from the main topics. These terms clarified the collective impact on improving health communication dynamics. The synthesis of the role of health IT in health communication models underscores its critical role in shaping patient-centered health care frameworks. This review highlights the significant contributions of key topics that should be thoroughly investigated and integrated into health communication models in the digital age. While health IT plays an essential role in promoting shared decision-making and patient empowerment, challenges such as usability, privacy concerns, and digital literacy remain significant barriers. Future research should prioritize evaluating these key themes and addressing the challenges associated with health IT in health communication models. Additionally, exploring how emerging technologies, such as artificial intelligence, can support these goals may provide valuable insights for enhancing health communication.

Analyzing resource saving and environmental safety in oil and gas waste processing system to provide sustainable development

The problems of environmental safety and resource conservation in the processing of industrial waste at Russian oil and gas enterprises are becoming more and more significant, especially in the context of current trends focused on decarbonization and sustainable development. The growth of industrial waste is not compensated by the corresponding growth in the volume and quality of processing, since only 10% of oily waste is processed, and the rest is buried or incinerated, which has an additional harmful impact on the environment. The article proposes a new approach to multifactor analysis of resource conservation and environmental safety in the integrated processing of industrial waste from oil and gas enterprises to achieve sustainable development goals. The algorithm developed by the authors, the theoretical basis of which is the method of data envelopment analysis (DEA), determines the sequence of solving interrelated problems of multifactor comparison of objects of the waste processing system by heterogeneous quality criteria: resource value, resource potential, resource saving and environmental safety. These problems are formulated on the basis of the super-efficiency model of the DEA method as a mathematical programming problem. The presented and tested approach to multifactor analysis of resource conservation and environmental safety of the waste treatment system allows: (1) to obtain and analyze relative comparative assessments of waste storage facilities and technologies regarding heterogeneous quality criteria, and (2) forming local and general decision-making strategies that provide sustainable development goals. The proposed approach appears promising for multiple prospective applications in the oil-bearing regions of the Russian Federation, where the growth in the formation of industrial waste has severe and irreversible environmental consequences.

Deep Reinforcemnet Learning for Robust Beamforming in Integrated Sensing, Communication and Power Transmission Systems.

A communication network integrating multiple modes can effectively support the sustainable development of next-generation wireless communications. Integrated sensing, communication, and power transfer (ISCPT) represents an emerging technological paradigm that not only facilitates information transmission but also enables environmental sensing and wireless power transfer. To achieve optimal beamforming in transmission, it is crucial to satisfy multiple constraints, including quality of service (QoS), radar sensing accuracy, and power transfer efficiency, while ensuring fundamental system performance. The presence of multiple parametric constraints makes the problem a non-convex optimization challenge, underscoring the need for a solution that balances low computational complexity with high precision. Additionally, the accuracy of channel state information (CSI) is pivotal in determining the achievable rate, as imperfect or incomplete CSI can significantly degrade system performance and beamforming efficiency. Deep reinforcement learning (DRL), a machine learning technique where an agent learns by interacting with its environment, offers a promising approach that can dynamically optimize system performance through adaptive decision-making strategies. In this paper, we propose a DRL-based ISCPT framework, which effectively manages complex environmental states and continuously adjusts variables related to sensing, communication, and energy harvesting to enhance overall system efficiency and reliability. The achievable rate upper bound can be inferred through robust, learnable beamforming in the ISCPT system. Our results demonstrate that DRL-based algorithms significantly improve resource allocation, power management, and information transmission, particularly in dynamic and uncertain environments with imperfect CSI.