Figure
Source publication
This paper provides a comprehensive state-of-the-art investigation of the recent advances in data science in emerging economic applications. The analysis is performed on the novel data science methods in four individual classes of deep learning models, hybrid deep learning models, hybrid machine learning, and ensemble models. Application domains in...
Similar publications
This paper provides the state of the art of data science in economics. Through a novel taxonomy of applications and methods advances in data science are investigated. The data science advances are investigated in three individual classes of deep learning models, ensemble models, and hybrid models. Application domains include stock market, marketing...
This paper provides a state-of-the-art investigation of advances in data science in emerging economic applications. The analysis was performed on novel data science methods in four individual classes of deep learning models, hybrid deep learning models, hybrid machine learning, and ensemble models. Application domains include a wide and diverse ran...
This paper provides the state of the art of data science in economics. Through a novel taxonomy of applications and methods advances in data science are investigated. The data science advances are investigated in three individual classes of deep learning models, ensemble models, and hybrid models. Application domains include stock market, marketing...
Citations
... GANs generate new data by learning very complex relationships and structures among different kinds of data, and they can generate large amounts of data that then feed a wide variety of deep learning models. (27,28) The discriminative model, which tries to distinguish between the fake and real data, is modeled by deep neural networks that are often referred to as the classifier. The generative model, modeled by deep neural networks, is used to produce 'fake' data. ...
This article explores the transformative role of artificial intelligence and machine learning in tackling climate change. It highlights how advanced computational techniques enhance our understanding and response to environmental shifts. Machine learning algorithms process vast climate datasets, revealing patterns that traditional methods might overlook. Deep learning neural networks, particularly effective in climate research, analyze satellite imagery, climate sensor data, and environmental indicators with unprecedented accuracy. Key applications include predictive modeling of climate change impacts. Using convolutional and recurrent neural networks, researchers generate high-resolution projections of temperature rises, sea-level changes, and extreme weather events with remarkable precision. AI also plays a vital role in data integration, synthesizing satellite observations, ground-based measurements, and historical records to create more reliable climate models. Additionally, deep learning algorithms enable real-time environmental monitoring, tracking changes like deforestation, ice cap melting, and ecosystem shifts. The article also highlights AI-powered optimization models in mitigation efforts. These models enhance carbon reduction strategies, optimize renewable energy use, and support sustainable urban planning. By leveraging machine learning, the research demonstrates how AI-driven approaches offer data-backed solutions for climate change mitigation and adaptation. These innovations provide practical strategies to address global environmental challenges effectively.
... This results in an improved accuracy and robustness. Studies for example, Nosratabadi et al. (2020) reports that integrating Bayesian methods with ML significantly enhances predictive accuracy power, while Linardatos et al. (2020) highlights that hybrid models improve interpretability and robustness in complex data environments. Thus, hybrid approaches represent a significant advancement in modelling approaches and methodologies. ...
Background: This study aimed to fill a critical research gap by comparing traditional Structural Equation Modelling (SEM) with hybrid Bayesian-Machine Learning (ML) models in marketing research, focusing on the limited exploration of these advanced techniques. Purpose: This study aimed to evaluate the effectiveness of integrating Bayesian SEM with advanced machine learning techniques to enhance predictive model performance, manage complex data structures, and improve marketing applications. Design/methodology/approach: The study employed a systematic comparative research design to assess the predictive accuracy and robustness of traditional SEM in comparison to hybrid Bayesian-(Bayesian-ML) models. A rigorous review of 262 scholarly articles from major databases was conducted, with 23 studies meeting inclusion criteria to inform the model development and evaluation. Findings/Result: The findings show that traditional SEM excels in theoretical modelling and interpretability but lacks predictive accuracy and robustness, which Bayesian SEM improves by using prior distributions. ML techniques further enhance predictive accuracy and robustness, while hybrid models combining Bayesian SEM with ML achieve the highest levels of both. Conclusion: Adopting hybrid models can substantially enhance the predictive accuracy of marketing outcomes and the robustness of model analyses. Originality/value (State of the art): This study contributes to knowledge by advancing methodological approaches through challenging existing data analysis paradigms, methods and approaches and therebefore offering practical guidance for future studies.
... MSE is a commonly used loss function in machine learning. It evaluates predictive performance by calculating the average squared errors between predicted and actual values [50]. Since MSE measures squared error, a larger squared error results in a higher MSE [51]. ...
... Since MSE measures squared error, a larger squared error results in a higher MSE [51]. Therefore, a lower MSE indicates better model performance [50]. ...
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.
... The use of machine learning tools to solve various mathematical and economic problems is increasingly being applied in modern educational practice. This is pointed out, for example, by Jordan & Mitchell 2015;Nosratabadi et al. 2020;Tedre et al. 2021, Zhong et al. 2021Mukhamediev et al. 2022;Tehranian, K. 2023 and other researchers. It is emphasized that the use of machine learning methods allows to expand the range and types of data processed and to perform analysis with higher speed, minimize errors in calculations, etc. ...
... Literature suggests that AI-based models can enhance battery management by predicting charge-discharge cycles, minimizing degradation, and improving costeffectiveness. By integrating AI into energy storage, renewable energy sources become more reliable and scalable [93][94][95]. ...
... However, despite the advantages, challenges such as data availability, computational complexity, regulatory barriers, and cybersecurity risks persist. AI-driven energy solutions require vast amounts of high-quality data to train predictive models, and limitations in data accessibility can affect performance [94][95][96][97][98][99][100][101][102][103][104][105][106][107][108][109][110]. Additionally, the high computational demands of AI algorithms necessitate substantial infrastructure and processing power, which may be a barrier to widespread adoption. ...
Renewable energy has emerged as a critical component in the global pursuit of sustainable development and carbon neutrality. However, the inherent challenges associated with renewable energy sources-such as intermittency, variability, and storage limitations-necessitate innovative solutions to enhance efficiency and reliability. The integration of Machine Learning (ML) and Artificial Intelligence (AI) has revolutionized the energy sector by optimizing renewable energy generation, forecasting demand, and improving grid stability. Data Science plays a pivotal role in processing vast amounts of energy-related data, enabling accurate predictions and data-driven decision-making. This paper explores how ML, AI, and Data Science contribute to advancements in renewable energy technologies, covering aspects such as predictive maintenance, smart grids, and energy storage optimization. A comprehensive literature review presents key research findings in the domain, demonstrating the application of AI and ML in energy management and predictive modeling. The research methodology section outlines the data-driven approaches used to optimize energy utilization, followed by an in-depth analysis of results obtained from AI-driven models. The study concludes with insights into future research directions, policy implications, and the potential of AI-augmented energy systems in fostering a more resilient and sustainable energy future. Machine Learning (ML) and Artificial Intelligence (AI) play a pivotal role in advancing renewable energy by leveraging data science to optimize energy generation, distribution, and consumption. Through predictive analytics, ML models enhance the efficiency of solar and wind power by forecasting energy output based on weather patterns, historical data, and real-time inputs. AI-driven algorithms improve grid stability by balancing supply and demand, reducing energy wastage, and integrating diverse renewable sources. Additionally, data science enables fault detection, predictive maintenance, and energy storage optimization, ensuring a more reliable and cost-effective renewable energy infrastructure. As AI and ML continue to evolve, their application in renewable energy promises a more sustainable and efficient future.
... There is growing interest in leveraging these data-driven methods to tackle previously insurmountable challenges in the modeling and prediction of dynamical systems across numerous applications. These include, but are not limited to, health sciences [6-10], biology [11,12], epidemiology [13,14], ecology and climate science [15][16][17][18][19], financial markets and economics [20][21][22], and engineering [23][24][25][26][27][28][29][30], among many others. ...
Human behavior presents significant challenges for data-driven approaches and machine learning, particularly in modeling the emergent and complex dynamics observed in social dilemmas. These challenges complicate the accurate prediction of strategic decision-making in structured populations, which is crucial for advancing our understanding of collective behavior. In this work, we introduce a novel approach to predicting high-dimensional collective behavior in structured populations engaged in social dilemmas. We propose a new feature extraction methodology, Topological Marginal Information Feature Extraction (TMIFE), which captures agent-level information over time. Leveraging TMIFE, we employ a graph neural network to encode networked dynamics and predict evolutionary outcomes under various social dilemma scenarios. Our approach is validated through numerical simulations and transfer learning, demonstrating its robustness and predictive accuracy. Furthermore, results from a Prisoner's Dilemma experiment involving human participants confirm that our method reliably predicts the macroscopic fraction of cooperation. These findings underscore the complexity of predicting high-dimensional behavior in structured populations and highlight the potential of graph-based machine learning techniques for this task.
... This multi-model synthesis has become a critical research focus in various fields [15]. Hybrid models, which combine different characteristics, are particularly suited for handling the complexity of wind power prediction data, enhancing accuracy, and significantly reducing errors [16]. To improve prediction accuracy, data-driven approaches and advanced deep learning algorithms have emerged as key areas of research interest [17,18]. ...
This paper proposes a dual-loop back-to-back converter coordination control scheme with a DC-side voltage as the primary control target, along with a CROW unloading control strategy for low voltage ride-through (LVRT) capability enhancement. The feasibility and effectiveness of the proposed system topology and control strategy are verified through MATLAB/Simulink simulations. Furthermore, a hybrid short-term wind power prediction model based on data-driven and deep learning techniques (CEEMDAN-CNN-Transformer-XGBoost) is introduced in the wind turbine control system. The coordination control strategy seamlessly integrates wind power prediction, pitch angle adjustment, and the control system, embodying a predictive-driven intelligent optimization control approach. This method significantly improves prediction accuracy and stability, theoretically reduces unnecessary pitch angle adjustments, lowers mechanical stress, and enhances system adaptability in complex operating conditions. The research findings provide a valuable theoretical foundation and technical reference for the intelligent and efficient operation of wind power generation systems.
... Generalized ordered logit relaxes the assumption and is more powerful than ordered logit (Kanyenji et al. 2020). In fact, the advantages of this model include flexibility in modeling with structured management and complex relationships, managing missing data, clear interpretation of parameters (Kleiner et al. 2021), accurate prediction of change trends (Nosratabadi et al. 2020), integration with other statistical methods, and application in various fields such as marketing, finance, healthcare, and social sciences (Kim et al. 2020). This model is as follows (Williams, 2006): ...
... The blockchain started with Satoshi Nakamoto's idea in 2008 in the Bitcoin whitepaper [1]. It changed digital transactions by suggesting a new way to do them without banks. ...
In the last ten years-blockchain technology has changed a lot. It started as basic records shared across many computers and has become advanced networks that can do many things. Therefore, this study looks at how well blockchain systems can handle complex math problems for security. We focus on creating a system that puts timestamps on events. This method combines Schnorr signatures and Pedersen commitments. We use the Ethereum network to store and check data. The system uses a smart contract to manage data and connect servers and clients. Servers put timestamps on events and clients send data to be timestamped and checked. This setup makes sure they are clear and reliable. However, the study tests how well the system works by looking at cost-how much memory it uses and how fast it runs. The study also talks about using Ethereum. Ethereum lets everyone see timestamps and what happen-which is good for trust. But it can't keep secrets-so it's not good for private data. This study shows that blockchain is useful for more than just money. It can make sure events are real and safe. By mixing math ideas in blockchain-the study shows how to make data safer and more reliable in many places.
... Machine learning, a term widely used across diverse fields, lacks a universally agreed-upon definition due to its broad applicability and the diverse contributions of researchers from various disciplines [1]- [3] This ambiguity is rooted in the extensive areas it covers and the collaborative efforts of researchers with diverse backgrounds. In a broad sense, machine learning can be understood as an algorithmic framework facilitating data analysis, inference, and the establishment of preliminary functional relationships. ...
... However, after pre-processing and feature engineering, the number of features increased to 3821. The machine learning model trained on this big dataset contains over three (3) thousand features [41]. Analysis results with confusion matrix are shown in fig. ...
Machine learning (ML) has become a ubiquitous tool across various domains of data mining and big data analysis. The efficacy of ML models depends heavily on high-quality datasets, which are often complicated by the presence of missing values. Consequently, the performance and generalization of ML models are at risk in the face of such datasets. This paper aims to examine the nuanced impact of missing values on ML workflows, including their types, causes, and consequences. Our analysis focuses on the challenges posed by missing values, including biased inferences, reduced predictive power, and increased computational burdens. The paper further explores strategies for handling missing values, including imputation techniques and removal strategies, and investigates how missing values affect model evaluation metrics and introduces complexities in cross-validation and model selection. The study employs case studies and real-world examples to illustrate the practical implications of addressing missing values. Finally, the discussion extends to future research directions, emphasizing the need for handling missing values ethically and transparently. The primary goal of this paper is to provide insights into the pervasive impact of missing values on ML models and guide practitioners toward effective strategies for achieving robust and reliable model outcomes.
