Bogdan Petru Vrinceanu

The rapid development of cryptocurrency markets has intensified the use of algorithmic trading systems capable of operating under high volatility and non-stationary market conditions. In this context, machine learning–based trading bots have gained increasing attention due to their potential to adapt dynamically to evolving price patterns. This study investigates the performance of two supervised machine learning classification approaches—k-Nearest Neighbors (kNN) and Learning Vector Quantization (LVQ)—when applied to automated trading in the Bitcoin market. The research addresses the following question: Do adaptive prototype-based learning models outperform instance-based classification models in algorithmic trading applications under identical market conditions? To answer this question, two fully automated trading bots were developed by the authors and implemented on the TradingView platform using Pine Script. Both bots were evaluated under identical experimental settings, trading the BTC/USDT pair on an hourly time frame over the period January 2025 to January 2026, with an initial capital of 1000 USD. Additional market strength and trend filters were applied to enhance signal robustness. The empirical results reveal notable differences between the two learning paradigms. The kNN-based trading bot exhibited a conservative trading behavior, characterized by lower trade frequency and moderate drawdown, but failed to generate positive cumulative returns. In contrast, the LVQ-based trading bot achieved superior profitability and demonstrated a stronger capacity to adapt to changing market regimes, albeit at the cost of significantly higher drawdown and increased risk exposure. The findings suggest that adaptive prototype-based learning models such as LVQ are better suited for algorithmic trading in highly volatile cryptocurrency markets when profitability and adaptability are prioritized. However, the increased risk associated with such models highlights the importance of incorporating robust risk management mechanisms. Overall, this study contributes to the literature by providing a controlled, implementation-level comparison of two machine learning trading strategies using realistic market conditions and fully automated trading systems.

Cryptocurrency markets are characterized by elevated volatility, structural instability, and rapidly changing investor sentiment, which significantly challenge traditional portfolio optimization methodologies. Under such conditions, static portfolio allocation models frequently fail to ade-quately incorporate uncertainty, behavioral adaptation, and dynamic market responsiveness into the investment decision-making process. In parallel, recent advances in artificial intelligence and data-driven financial systems have transformed modern portfolio management by enabling adaptive investment strategies capable of processing large volumes of financial information in real time. Nevertheless, AI-assisted portfolio systems often remain highly sensitive to noisy market signals, parameter instability, and excessive portfolio rebalancing, particularly in volatile crypto-currency environments. This study proposes a behaviorally adaptive portfolio optimization framework under interval uncertainty that integrates robust optimization principles, behavioral finance mechanisms, non-linear transaction costs, and AI-assisted allocation adjustment within a unified cryptocurrency in-vestment structure. The proposed methodology introduces a behavioral preference parameter capable of dynamically adjusting portfolio composition according to varying investor attitudes toward uncertainty and market expectations. Simultaneously, interval uncertainty modeling is employed to represent ambiguous financial parameters through bounded intervals, while artificial intelligence mechanisms act as adaptive decision-support tools that improve portfolio respon-siveness under changing market conditions. The empirical analysis is conducted using major cryptocurrency assets, including Bitcoin, Ethereum, Solana, and Binance Coin, over the period January–June 2025. Multiple behavioral portfolio configurations are evaluated in order to analyze the interaction between profitability, downside risk, portfolio stability, and transaction cost efficiency. The results indicate that the proposed framework improves risk-adjusted portfolio performance and generates more stable allocation strategies compared to traditional static portfolio optimization approaches. In particular, the integration of behavioral adaptation and robust optimization contributes to reduced sensitivity to estimation uncertainty and unstable market fluctuations. The findings further demonstrate that AI-assisted adaptive allocation mechanisms improve portfolio flexibility and support smoother portfolio transitions under volatile market conditions. Moreover, the incorporation of nonlinear transaction costs leads to more realistic and practically implementable cryptocurrency investment strategies. Overall, the proposed framework provides a robust and behaviorally interpretable approach for modern cryptocurrency portfolio management and contributes to the growing literature at the intersection of artificial intelligence, behavioral fi-nance, and uncertainty-aware portfolio optimization.

Modern financial markets are increasingly shaped by algorithmic trading systems and artificial intelligence techniques that process large volumes of financial data in real time. However, machine learning-based trading systems often suffer from signal instability and excessive sensitivity to market noise, which may lead to overtrading and increased financial risk. In highly volatile environments such as cryptocurrency markets, the reliability of trading signals becomes a critical issue for both portfolio allocation and risk management. This study proposes an entropy-filtered machine learning framework designed to enhance the stability and risk-awareness of algorithmic trading strategies. The proposed approach integrates entropy-based filtering techniques with machine learning classifiers in order to reduce noise in market signals, thereby improving the risk-adjusted stability of trading strategies. Entropy measures are employed as a filtering mechanism that evaluates the informational content of predictive signals and suppresses unreliable model outputs. The empirical analysis is conducted using cryptocurrency market data, where the entropy-filtered framework is applied to trading signal generation and decision making. The results indicate that the proposed approach improves the stability of trading signals and reduces the occurrence of false signals compared to conventional machine learning models. In addition, entropy filtering contributes to a more balanced risk–return profile and enhances the overall robustness of trading strategies. Moreover, entropy filtering contributes to a more balanced risk–return profile and enhances the overall robustness of trading strategies. The findings suggest that entropy-based filtering substantially improves the reliability and risk-awareness of machine learning trading systems, providing a promising direction for the development of more robust AI-driven financial decision frameworks.

Modern financial markets are increasingly shaped by algorithmic trading systems and artificial intelligence techniques that process large volumes of financial data in real time. However, machine learning–based trading systems often suffer from signal instability and excessive sensitivity to market noise, which may lead to overtrading and increased financial risk. In highly volatile environments such as cryptocurrency markets, the re-liability of trading signals becomes a critical issue for both portfolio allocation and risk management. This study proposes an entropy-filtered machine learning framework designed to en-hance the stability and risk-awareness of algorithmic trading strategies. The proposed approach integrates entropy-based filtering techniques with machine learning classifiers in order to reduce noise in market signals improving the risk-adjusted stability of algo-rithmic trading strategies. Entropy measures are employed as a filtering mechanism that evaluates the informational content of market signals and suppresses unreliable predic-tions generated by the learning model. The empirical analysis is conducted using cryp-tocurrency market data, where the entropy-filtered machine learning framework is ap-plied to trading signal generation and portfolio decision making. The results indicate that the proposed approach improves the stability of trading signals and reduces the occur-rence of false signals compared to conventional machine learning trading models. Moreover, the integration of entropy filtering contributes to a more balanced risk–return profile and enhances the overall robustness of algorithmic trading strategies.The findings suggest that combining information-theoretic measures with machine learning tech-niques represents a promising direction for developing more reliable and risk-aware financial decision systems. The results suggest that entropy-based filtering can substan-tially improve the robustness and risk-awareness of machine learning trading systems, providing a promising direction for future AI-driven financial decision frameworks.

This study investigates post-crisis digital transformation and business model adaptation in Eastern Europe and the Middle East, regions with diverse economic structures and varying technological maturity. It explores how organizations in these regions utilized digital technologies to navigate crisis conditions, focusing on the interplay between digital agility, data-driven strategies, and business sustainability. Existing literature emphasizes the growing importance of digital transformation for organizational resilience, particularly during disruptions. However, the specific challenges and opportunities within emerging economies like those in Eastern Europe and the Middle East remain underexplored. This research addresses this gap by examining how digital agility influences business sustainability post-crisis, whether digital transformation effectiveness varies across sectors, and what key drivers and barriers exist for successful digital transformation in these regions. Key findings reveal a strong positive correlation between digital agility and post-crisis business sustainability; significant sectoral variations in digital transformation success, with financial services and healthcare leading; regional differences in digital adoption and effectiveness, with Middle Eastern companies demonstrating higher efficiency gains despite structural barriers; the importance of integrating digital strategies beyond operational efficiency for sustained growth; and crucial roles for leadership, digital governance, employee training, external partnerships, cloud computing, and AI, while common barriers include implementation costs, workforce resistance, and cybersecurity concerns. These findings contribute empirical evidence of the link between digital transformation and post-crisis resilience in emerging economies. The study highlights the importance of a holistic approach, encompassing not only technology but also strategic alignment, leadership, human capital, and customer-centricity. It underscores the need for enabling policy environments, including infrastructure investment, SME support, and robust cybersecurity frameworks. Limitations include geographical focus and sample size. Future research could explore long-term digital sustainability, emerging technology impacts, and sector-specific challenges.

This research investigates the transformative influence of Fintech on reshaping business models within the dynamic context of Eastern Europe and the Middle East (EEME) in the post-COVID-19 era. The study examines how the pandemic has accelerated Fintech adoption and its subsequent impact on business model evolution across various sectors in the region. Existing literature highlights the disruptive potential of Fintech globally, but there is a gap in understanding its specific impact on the EEME region, particularly in the post-pandemic recovery phase. This study addresses this gap by exploring the complex interplay between Fintech adoption and business model transformation. A large-scale survey of 1050 businesses operating in various sectors across EEME was conducted. The survey captured data on Fintech adoption rates, types of solutions utilized, changes in key aspects of business models (revenue generation, customer acquisition, operations, value proposition), enabling factors (regulatory support, consumer demand, digital infrastructure), and perceived challenges (cybersecurity, data privacy). Statistical analysis, including correlation and regression analysis, was employed to test the research hypotheses. The findings reveal a significant acceleration of Fintech adoption post-COVID-19, with 78% of businesses adopting new Fintech solutions compared to 32% pre-pandemic. Key drivers of adoption include regulatory support for digital finance (β = 0.62, p < 0.001), consumer demand for digital services (β = 0.55, p < 0.001), and digital infrastructure development (β = 0.48, p < 0.001). Sectoral variations in impact were observed, with finance (85% reporting significant changes) and retail (72%) experiencing greater transformation than manufacturing (48%). While cybersecurity and data privacy concerns were prevalent (68% and 55% respectively), their negative correlation with business model transformation was relatively weak (β = -0.25 and β = -0.20, p < 0.05). These results suggest that while challenges exist, they do not necessarily prevent businesses from adapting their models using Fintech. This study contributes to the field by providing empirical evidence of the transformative role of Fintech in EEME post-COVID-19, highlighting the key drivers and challenges, and revealing sectoral variations in impact. The findings have practical implications for businesses, policymakers, and investors in the region. The research acknowledges limitations inherent in survey-based methodologies, such as potential response bias and the cross-sectional nature of the data, which limits causal inferences. Future research could explore these dynamics through longitudinal studies and qualitative investigations.

This paper proposes a novel entropy-enhanced machine learning framework for algorithmic trading, designed to improve decision reliability and model interpretability under conditions of financial uncertainty. The core methodology integrates a Learning Vector Quantization (LVQ) classifier with a Shannon entropy-based filter that selectively activates trading signals based on their informational certainty. Mathematically, the system is formulated as a composite decision function constrained by entropy: only predictions satisfying H(x_t ) ≤θ are executed, where H(x_t ) is the Shannon entropy of the class probability distribution at time t. The empirical evaluation spans three asset classes—cryptocurrencies, foreign exchange, and equities—demonstrating significant improvements in trade precision, Sharpe ratio, and robustness to market volatility. The entropy filter acts as a regularization mechanism, reducing overfitting and mitigating the propagation of noisy signals. Beyond its empirical success, the model contributes a formalized framework for entropy-regularized decision making, opening avenues for generalized entropic models such as Tsallis and Kaniadakis entropy in future research. This entropy-aware structure bridges statistical learning theory and real-world financial applications by introducing a quantifiable measure of confidence into the machine learning pipeline. The results underscore the model’s potential as a robust, interpretable, and adaptive solution for algorithmic trading in complex and uncertain markets.

Portfolio optimization is a fundamental problem in financial theory, aiming to balance risk and return in asset allocation. Traditional models, such as Mean–Variance optimization, are effective, but often fail to account for diversification adequately. This study introduces the Mean–Variance–Entropy (MVE) model, which integrates Tsallis entropy into the classic Mean–Variance framework to enhance portfolio diversification and risk management. Entropy, specifically second-order entropy, penalizes excessive concentration in the portfolio, encouraging a more balanced and diversified allocation of assets. The model is applied to a portfolio of five major cryptocurrencies: Bitcoin (BTC), Ethereum (ETH), Solana (SOL), Cardano (ADA), and Binance Coin (BNB). The performance of the MVE model is compared with that of the traditional Mean–Variance model, and results demonstrate that the entropy-enhanced model provides better diversification, although with a slightly lower Sharpe ratio. The findings suggest that while the entropy-adjusted model results in a slightly lower Sharpe ratio, it offers better diversification and a more resilient portfolio, especially in volatile markets. This study demonstrates the potential of incorporating entropy into portfolio optimization as a means to mitigate concentration risk and improve portfolio performance. The approach is particularly beneficial for markets such as cryptocurrency, where volatility and asset correlations fluctuate rapidly. This paper contributes to the growing body of literature on portfolio optimization by offering a more diversified, robust, and risk-adjusted approach to asset allocation

The financial market is inherently a complex and dynamic environment, where investors must balance the trade-off between expected return and associated risk. While the clas-sical Markowitz mean-variance model has served as a cornerstone for modern portfolio theory, it may not fully capture the informational uncertainty present in asset returns. Recent studies suggest that entropy-based measures can enhance the robustness of portfolio optimization by accounting for the uncertainty embedded in return distributions. This paper proposes an extended mean-variance-entropy framework for portfolio selection, aiming to maximize a return-to-risk adjusted with entropy ratio. A mathematical formulation of the optimization problem is developed, and an analytical solution for the optimal asset weights is derived. Furthermore, a practical application is conducted using two assets, highlighting the potential benefits of incorporating entropy in portfolio decision-making.

This study explores the use of Shannon entropy as a filtering mechanism to enhance the trading signals produced by the LVQ (Learning Vector Quantization) machine learning algorithm in algorithmic trading. The integration of Shannon entropy aims to improve trade entry accuracy by reducing market noise and identifying stronger trends. A trading bot was developed and tested on Bitcoin using a three-minute timeframe on the TradingView platform, with backtesting conducted from February 1 to February 18, 2025. The fully automated strategy used 100% of available capital per trade, reinvesting profits for compounding. Positions were closed when an opposite signal was generated. A comparative analysis revealed that incorporating Shannon entropy outperformed a baseline strategy. These findings demonstrate that entropy-based filtering improves trade selection and profitability by reducing market noise and focusing on reliable trends, suggesting its potential for broader application in algorithmic trading.