A system design built on blockchain technology presents a fundamental challenge: the inherent transparency of the blockchain conflicts with the growing need for user privacy. This dissertation explores how Zero-Knowledge Proofs (ZKPs) can be strategically combined with blockchain to strike a balance between these competing demands. The dissertation analyzes the challenges of privacy and transparency and provides an overview of solutions across the privacy-transparency solution space, drawing from the author’s original research and the broader academic landscape. On the privacy-centric side, it proposes a privacy-preserving design that utilizes off-chain ZKPs. In contrast, on the transparency-centric side, it presents a transparency-enhancing design that leverages on-chain data to build trust. In the middle, it discusses a taxonomy of hybrid applications whose unique combination of on-chain privacy (via ZKPs) and blockchain transparency reveals both a disruptive potential and significant regulatory challenges.
The study proposes a blockchain-based framework to overcome the challenges of data silos, privacy risks, and interoperability limitations in pharmacovigilance systems, focusing on the refinement of adverse drug reaction (ADR) data collection, storage, and analysis. Incorporating blockchain's transparency, immutability, and security, the framework comprises four core components: a data collection layer for multi-source ADR reporting, a standardization layer for data integration and validation, a blockchain network layer for tamper-proof storage and secure sharing, and a data analysis layer for real-time risk detection and visualization. The framework's efficacy in drug safety monitoring and regulatory efficiency is exemplified by the MediLedger and Merck's SAP Pharma Blockchain Proof of Concept case studies. The proposed resolution shows remarkable potential for advancing global pharmacovigilance practices.
Modern cyber threats, known for their complexity and constant change, surpass traditional intrusion detection systems (IDS). This paper explores a new security approach that combines Artificial Intelligence (AI) with decentralized architectures to develop IDS that are robust, scalable, and protect user privacy. It examines the core roles of Federated Learning (FL) and Blockchain, highlighting three main research challenges: The vulnerability of AI models to adversarial attacks, privacy and data integrity concerns in collaborative learning, and performance limitations in distributed systems. To address these issues, we suggest solutions such as adversarial training, differential privacy, and lightweight consensus mechanisms. Our analysis of case studies shows that hybrid FL-Blockchain systems outperform traditional methods in practical application environments.
Oleg Kravets, Ali Husein, Diana Getmanskaia, Mustafa Al-Imari · 7 authors
The article discussed an automata model of the functioning process of a system with distributed ledger technology based on a blockchain. The goal is to develop a special mathematical system with a distributed ledger based on a blockchain, taking into account the specifics of implementing algorithms for mutual information coordination, the possibility of combining individual nodes into groups, and implementing alternative strategies based on the development of appropriate models and algorithms that ensure increased stability in their operation. Automata theory was used, theory and methods of mutual coordination were applied, and the mechanism of the functioning process of a system with a distributed register was implemented. An automata model has been obtained and investigated. Thus, an automata model of the distributed ledger technology blockchain system node functioning process has been developed, which differs from the node representation by a finite state machine with a variable structure and a linear tactic with the possibility of implementing non-standard functions: the formation of a branch of processed data and a temporary blockage attack, and provides for obtaining the dependence of the sequence of changing the node's behaviour strategy options on the conditions of the environment it interacts with.
A system implementaion paper on a decentralized application for transparent charity transactions built on hardhat , next js ,and django that combines the web2 and web3 dynamics.
We examine the qualifying attributes of decentralized finance (DeFi) as a financial asset class. To achieve this objective, we perform analysis on the relationship (using both level and percentage-change data) between DeFi valuation and selected influencing variables, namely total value locked (TVL), Bitcoin (BTC) value, and market variables. A suite of long-panel data econometric methods is employed on a multi-frequency (daily, weekly, and monthly) panel dataset comprising 16 major DeFi protocols from January 2022 to December 2023. Our empirical design aims to be a comprehensive assessment and triangulation. There are several key findings. First, while there is evidence of cointegration suggesting a possible long-run relationship, this relationship is found to be inconsistent across different variables and time frequencies. However, the impulse response analysis suggests that shocks from the influencing variables do not have a permanent impact. Second, Bitcoin value is found to be the most important influencing factor (positive and highly significant), reflecting strong cryptocurrency market sentiment and aligning with previous research on spillover effects from major cryptocurrencies (Șoiman et al., 2022; Yousaf et al., 2022).
Το Web3 δεν αποτελεί απλώς μια τεχνολογική εξέλιξη, αλλά μια ριζική μετατόπιση με βαθιές φιλοσοφικές προεκτάσεις. Οραματίζεται ένα Διαδίκτυο όπου οι χρήστες βρίσκονται στο επίκεντρο, χωρίς την ανάγκη ύπαρξης κεντρικής αρχής. Στόχος του είναι να διασφαλίσει ότι οι χρήστες έχουν την πλήρη ιδιοκτησία τόσο των δεδομένων που παράγουν όσο και της αξίας — οικονομικής ή πληροφοριακής — που προκύπτει από αυτά. Αυτή η μετάβαση σε ένα νέο τεχνολογικό υπόδειγμα βρίσκει εφαρμογή σε ποικίλους τομείς, όπως η Αποκεντρωμένη Αποθήκευση (Decentralized Storage), η οποία επιτρέπει ασφαλείς, κατανεμημένες λύσεις αποθήκευσης δεδομένων με αυξημένη ανθεκτικότητα στη λογοκρισία· τα Μη Ανταλλάξιμα Διακριτικά (Non-Fungible Tokens – NFTs), που εγγυώνται την ιδιοκτησία και την αυθεντικότητα ψηφιακών περιουσιακών στοιχείων· και τα Αποκεντρωμένα Παιχνίδια (Decentralized Gaming), τα οποία αξιοποιούν την τεχνολογία blockchain για να δημιουργήσουν οικονομίες που ανήκουν στους παίκτες, αποδεικτικά σπάνια ψηφιακά αγαθά και διαφανείς μηχανισμούς παιχνιδιού. Μαζί με πλήθος άλλων καινοτομιών, οι παραπάνω τεχνολογίες διαμορφώνουν τη νέα εποχή του Διαδικτύου. Όπως είναι αναμενόμενο, το Web3 έχει προσελκύσει το ενδιαφέρον της ερευνητικής κοινότητας, η οποία προσπαθεί να το θεμελιώσει εκ νέου, βασιζόμενη σε αναδυόμενες και ακόμη ανώριμες τεχνολογίες. Παράλληλα όμως, έχει κινήσει και το ενδιαφέρον κακόβουλων παραγόντων, που εκμεταλλεύονται τον πρώιμο χαρακτήρα και την πολυπλοκότητα αυτών των αλληλένδετων συστημάτων προς ίδιον όφελος. Η πρόκληση, επομένως, είναι να εξασφαλιστεί ότι η ασφάλεια θα εξελίσσεται παράλληλα με την ανάπτυξη του οικοσυστήματος του Web3, ώστε να μην εξελιχθεί σε ένα ασταθές ή εχθρικό περιβάλλον. Η συμβολή της παρούσας διατριβής σε αυτήν την προσπάθεια είναι πολυδιάστατη. Αρχικά, μελετούμε τη σχετική βιβλιογραφία σχετικά με την αλυσίδα συστοιχιών (blockchain) του Ethereum, τα NFTs και το Interplanetary File System (IPFS), το οποίο αποτελεί θεμελιώδες στοιχείο του επιπέδου αποθήκευσης δεδομένων του Web3, με στόχο τον εντοπισμό ευπαθειών και την ανάλυση του βαθμού ύπαρξης κακόβουλης δραστηριότητας. Στη συνέχεια, υιοθετώντας την οπτική των κακόβουλων χρηστών, εξετάζουμε πιθανούς τρόπους εκμετάλλευσης των παραπάνω τεχνολογιών και τεκμηριώνουμε πιθανούς διαύλους επίθεσης, ώστε να είναι ευκολότερος ο εντοπισμός και η αντιμετώπισή τους. Τέλος, προτείνουμε βελτιώσεις στον σχεδιασμό κρίσιμων υπηρεσιών του επιπέδου εφαρμογών του Web3, οι οποίες ενισχύουν τη διαθεσιμότητα και την επεκτασιμότητά τους, θέτοντας έτσι τα θεμέλια για πιο ανθεκτικές, επεκτάσιμες και μελλοντικά βιώσιμες αποκεντρωμένες εφαρμογές.
Social sustainability in urban and architectural design depends on inclusive, participatory processes that empower communities to actively engage in shaping their environments. This study investigates how emerging digital platforms, specifically Augmented Reality (AR) and decentralized platforms built on blockchain technology (Web3), can function as instruments for broadening public participation and enhancing perceptual access to urban art proposals. An original algorithm generated nine digital abstract sculptures, each with descriptive attributes forming the basis for qualitative analysis across different visualization modes: traditional renderings, Augmented Reality environments, and NFT-based Web3 representations. Through participant voting, each digital sculpture accumulated a measurable level of preference that served to identify which sculpture was perceived as most successful within each visualization context. Comparative analysis revealed how distinct digital interactions shape perception, engagement, and inclusivity of feedback processes. Regression models further predicted voting outcomes, showing that different sculptural attributes played a dominant role depending on the type of visualization. Findings indicate that platform-specific technological affordances substantially shape participatory outcomes. Consequently, the study argues that careful analysis and selection of the digital platform must precede any participatory process, as platform-specific affordances fundamentally condition the inclusivity, accessibility, and overall effectiveness of public engagement in socially sustainable design.
This paper introduces Lean 5.0, a human-centric evolution of Lean-Digital integration that connects predictive analytics, AI collaboration, and continuous learning within Industry 5.0 and Construction 5.0 contexts. A systematic literature review (2019-2024) and a 12-week empirical validation study demonstrate measurable performance gains, including a 13% increase in Plan Percent Complete (PPC), 22% reduction in rework, and 42% improvement in forecast accuracy. The study adopts a mixed-method Design Science Research (DSR) approach aligned with PRISMA 2020 guidelines. The paper also examines integration with digital twin and blockchain technologies to improve traceability, auditability, and lifecycle transparency. Despite limitations related to sample size, single-case design, and study duration, the findings show that Lean 5.0 provides a transformative paradigm connecting human cognition with predictive control in construction management.
Opening up data produced by the Internet of Things (IoT) and mobile devices for public utilization can maximize their economic value. Challenges remain in the trustworthiness of the data sources and the security of the trading process, particularly when there is no trust between the data providers and consumers. In this paper, we propose DEXO, a decentralized data exchange mechanism that facilitates secure and fair data exchange between data consumers and distributed IoT/mobile data providers at scale, allowing the consumer to verify the data generation process and the providers to be compensated for providing authentic data, with correctness guarantees from the exchange platform. To realize this, DEXO extends the decentralized oracle network model that has been successful in the blockchain applications domain to incorporate novel hardware-cryptographic co-design that harmonizes trusted execution environment, secret sharing, and smart contract-assisted fair exchange. For the first time, DEXO ensures end-to-end data confidentiality, source verifiability, and fairness of the exchange process with strong resilience against participant collusion. We implemented a prototype of the DEXO system to demonstrate feasibility. The evaluation shows a moderate deployment cost and significantly improved blockchain operation efficiency compared to a popular data exchange mechanism.
Andreea Elena Drăgnoiu, Andrei Ciobanu, Ruxandra F. Olimid
Self-Sovereign Identity (SSI) grants holders full ownership and control of their digital identities, being the ultimate digital identity model. Operating in a decentralized manner, SSI enables the verification of claims, including privacy-preserving mechanisms. Blockchain, which can be used to implement a Verifiable Data Registry (VDR), is often considered one of the pillars of SSI, along with Decentralized Identifiers (DIDs) and Verifiable Credentials (VCs). Unfortunately, blockchains are mostly siloed, affecting the interoperability and universality of SSI. We investigate the effect of blockchain isolation on blockchain-based SSI. We first define possible scenarios for cross-chain SSI and exemplify with real-life use cases. We then define specific requirements for cross-chain SSI and identify challenges, also in relation to the identified scenarios. We explore various solutions to achieve blockchain interoperability, with a focus on SSI. In particular, we identify the advantages and disadvantages of distinct cross-chain models for cross-chain SSI. Finally, we address the usability of cross-chain SSI and discuss security and privacy aspects, opening the way for future research.
Entrepreneurs typically seek financing in decentralized markets, where they approach investors sequentially. We develop a model of sequential capital markets with privately informed investors. The sequential market creates a dynamic adverse selection externality that leads to overinvestment and excessive rents to intermediaries, even as the number of competing investors becomes arbitrary large. The resulting rents lead to excessive entry of investors and insufficient entry of entrepreneurs. Moving to a centralized market structure or reducing transparency restores competitiveness but may harm efficiency. The model also explains how even a small skill advantage for an investor can lead to preferential deal flow and outsized returns.
This Master's thesis investigates the application of machine learning methods to cryptocurrency market prediction and the development of hybrid trading strategies that combine predictive signals with decentralized finance yield components. The study addresses how machine learning models can predict directional shifts in cryptocurrency markets and whether integrating DeFi yield elements can improve risk-adjusted portfolio returns compared to traditional buy-and-hold approaches. The empirical investigation examined multiple machine learning architectures for binary directional forecasting of Bitcoin price movements. Models were trained on data spanning January 2018 to August 2024 using walk-forward validation. LightGBMRegressor achieved 53 % directional accuracy, while Random Forest reached 52 % accuracy. Other tested models, including LSTM networks and MLP, performed within the 51-56 % accuracy range. These results indicate that while machine learning methods demonstrate potential for market direction prediction when combined with properly formatted datasets and appropriate technical indicators, achieving high prediction accuracy remains challenging. A composed trading strategy was developed that integrated LSTM predictions with real-world DeFi yield rates from liquidity pools. The strategy utilized actual yield data to provide realistic performance assessment. Despite modest directional prediction accuracy of 53 %, the hybrid approach reduced drawdown by 50 % compared to the benchmark buy-and-hold strategy. The DeFi yield component compensated for imperfect directional signals, demonstrating that yield-enhanced strategies can achieve adequate risk-adjusted returns even without superior prediction accuracy. The study also examined structural differences between decentralized and traditional financial systems. DeFi offers global accessibility, programmable infrastructure, and fast settlement, but faces challenges including security vulnerabilities and regulatory uncertainty. However, the primary contribution lies in demonstrating that hybrid strategies combining machine learning signals with DeFi yield mechanisms represent a viable approach to portfolio management, when effective risk management is implemented.
This study examines the adoption of Digital Ledger Technology (DLT) and its impact on the accuracy of financial reporting and the efficiency of auditing processes within Jordanian organizations. Using a quantitative research design, the study assesses how DLT enhances financial data integrity and supports real-time auditing capabilities. Data were collected from 210 accounting and auditing professionals representing five major Jordanian institutions: the Central Bank of Jordan, Jordan Customs Department, Arab Bank, Deloitte Jordan, and Ernst & Young Jordan. A structured questionnaire served as the primary data collection instrument, employing a five-point Likert scale to measure perceptions across key constructs related to DLT adoption. To improve response rates, the questionnaire was distributed through both physical and digital channels. Descriptive statistics were used to analyze demographic data, while multiple regression and independent samples t-tests were applied to test the study’s hypotheses. The results revealed a statistically significant and positive relationship between DLT adoption and both the accuracy and reliability of financial reporting, as well as between DLT utilization and enhanced auditing speed and efficiency. Regression analysis indicated that DLT adoption accounted for 52% of the variance in financial reporting accuracy, while t-test results confirmed significant differences between DLT-based and traditional auditing methods. The study complied with ethical standards, ensuring confidentiality and voluntary participation. Overall, the findings demonstrate that DLT plays a transformative role in improving accounting and auditing practices within a developing economy context.
Purpose This study analyzes Georgia’s healthcare transformation from 1991 through both historical policy evolution and the WHO Health Systems Framework, evaluating how reforms prioritized or neglected system components while projecting 2030 outcomes. Design/methodology/approach We analyzed 26 peer-reviewed studies, WHO reports and Georgian government records, categorizing reforms by WHO component and assessing their cumulative impact on equity and efficiency. Findings Georgia’s journey reveals three distinct phases: post-Soviet collapse (1991–2006), market-driven reforms (2007–2012) and universal coverage expansion (2013–present). While financing and service delivery improved (OOP payments reduced from 74.7 to 57%), chronic workforce shortages (1:1 nurse–doctor ratio) and governance gaps persist. The 2030 strategy represents the first holistic attempt to address all WHO components simultaneously. Practical implications Policymakers must balance historical lessons (e.g. corruption risks in decentralization) with WHO-aligned investments, particularly in workforce training and digital infrastructure, to achieve 2030 goals. Originality/value This is the first study to combine historical policy analysis with WHO framework application for Georgia, revealing how reform sequencing explains current strengths (90% coverage) and weaknesses (rural disparities).
One-Time Passwords (OTPs) are a core component of multi-factor authentication in banking, e-commerce, and digital platforms. However, conventional delivery channels such as SMS and email are increasingly vulnerable to SIM-swap fraud, phishing, spoofing, and session hijacking. This study proposes an end-to-end mobile authentication architecture that integrates a permissioned Hyperledger Fabric blockchain for tamper-evident identity management, an AI-driven risk engine for behavioral and SIM-swap anomaly detection, Zero-Knowledge Proofs (ZKPs) for privacy-preserving verification, and geolocation-bound OTP validation for contextual assurance. Hyperledger Fabric is selected for its permissioned governance, configurable endorsement policies, and deterministic chaincode execution, which together support regulatory compliance and high throughput without the overhead of cryptocurrency. The system is implemented as a set of modular microservices that combine encrypted off-chain storage with on-chain hash references and smart-contract–enforced policies for geofencing and privacy protection. Experimental results show sub-0.5 s total verification latency (including ZKP overhead), approximately 850 transactions per second throughput under an OR-endorsement policy, and an F1-score of 0.88 for SIM-swap detection. Collectively, these findings demonstrate a scalable, privacy-centric, and interoperable solution that strengthens OTP-based authentication while preserving user confidentiality, operational transparency, and regulatory compliance across mobile network operators.
T J E N N I N G, Kalokhe Omkar Nanabhau, Takale Ram Arjun, Borge Akash Sandip
The proliferation of digital documents and academic credentials in today's interconnected world has created both opportunities and vulnerabilities. Traditional certificate issuance and storage mechanisms are highly susceptible to forgery, duplication, and unauthorized manipulation, undermining the trustworthiness of academic and professional qualifications. To address these challenges, this research proposes a blockchain-based certificate generation and verification system that ensures transparency, immutability, and trust across stakeholders. Leveraging distributed ledger technology, the system securely records certificate metadata and unique identifiers, enabling real-time, tamper-proof validation without reliance on intermediaries. The architecture integrates modern web technologies such as Next.js for frontend and backend services, MongoDB for scalable storage, JWT for authentication, and cryptographic techniques including bcrypt for enhanced security. Additionally, smart contracts deployed on Ethereum/Ganache enable decentralized storage and validation, while certificate data is simultaneously linked with non-fungible tokens (NFTs) to provide verifiable ownership and authenticity. This integration not only eliminates certificate fraud but also facilitates seamless verification across institutions, employers, and regulatory authorities. By combining blockchain's decentralized security with user-friendly web applications, the proposed approach aims to create a globally interoperable, cost-effective, and future-ready framework for academic and professional certification systems.
This research aims to analyze the readiness of Indonesia’s legal framework to accommodate Non-Fungible Tokens (NFTs) as fiduciary collateral objects and to identify the potential obstacles in their execution process. NFTs are digital assets based on blockchain technology that possess unique characteristics and economic value, theoretically fulfilling the criteria of intangible objects under Law Number 42 of 1999 concerning Fiduciary Collateral. However, the current legal instruments in Indonesia have not yet explicitly recognized NFTs as eligible fiduciary objects. This study employs a normative juridical method with both statutory and case approaches, supported by secondary data derived from an interview with the Direktorat Jenderal Administrasi Hukum Umum (Ditjen AHU) to enrich the legal analysis. The findings indicate that Indonesia’s regulatory readiness regarding NFTs as fiduciary collateral remains conceptual rather than operational. The absence of a registration mechanism and valuation system for digital assets creates significant legal uncertainty for both creditors and debtors. Furthermore, the execution of NFTs presents additional challenges, including the dependence on private keys, the incompatibility of automatic smart contract transfers with the due process of law principle, and the speculative volatility of NFT market values. Therefore, the study suggests the need for new legal norms and an authorized digital asset valuation institution to ensure that NFTs can be effectively and lawfully integrated into Indonesia’s fiduciary security system. Non-Fungible Token, Fiduciary Collateral, Collateral Execution, Intangible Assets
CONTEXT: Accurate preoperative prediction of occult lymph node metastasis (OLNM) in clinically lymph node negative (cN0) papillary thyroid carcinoma (PTC) is critical for optimizing therapeutic strategy, particularly for thermal ablation and active surveillance. OBJECTIVE: The aim of this study was to develop an interpretable machine-learning (ML) model to predict the risk of OLNM in cN0 PTC patients. METHODS: This retrospective study analyzed data of 961 cN0 PTC patients (August 2018-August 2023). Multivariable logistic regression identified independent risk factors for OLNM in cN0 PTC. The cohort was randomly divided into the training and test sets, and a subset of patients with tumors sized 1 cm or less was further extracted from the test set for internal validation. Eight ML models incorporating clinical, ultrasonographic, and molecular features were developed and evaluated. Shapley Additive exPlanations (SHAP) enhanced interpretability. RESULTS: RET fusion positivity and BRAF mutation positivity were identified as independent molecular risk factors for OLNM in cN0 PTC, alongside 6 clinical and ultrasonographic variables. Nine predictors were incorporated into the predictive model. The random forest (RF) model achieved optimal performance with an area under the curve (AUC) of 0.906 in the training set and 0.733 in the test set, along with the lowest Brier scores of 0.135 and 0.212, respectively. Analysis of tumors sized 1 cm or less internally validated the model's robustness with an AUC of 0.719. SHAP analysis identified size, age, and clustered punctate echogenic foci as the top predictors. CONCLUSION: This is the first study to identify RET fusion positivity as an independent OLNM risk factor in cN0 PTC. The developed RF model demonstrates moderate predictive performance for OLNM risk and provides a framework for integrating clinical, sonographic, and molecular data, and is deployed as a web calculator (https://predictingoccultlymphnodemetastasis.shinyapps.io/web3/).
Thyroid Cancer Diagnosis and Treatment
Artificial Intelligence in Healthcare and Education
This study presents a systematic literature review (SLR) conducted under the PRISMA 2020framework to investigate the convergence of two transformative paradigms: Generative ArtificialIntelligence (GenAI) and Web3. The findings indicate that, while each technology independentlydrives digital transformation, their integration remains underexplored. GenAI advancesinnovation through algorithmic creativity, personalization, and automated content generation,whereas Web3, enabled by blockchain, smart contracts, non-fungible tokens (NFTs), anddecentralized autonomous organizations (DAOs), introduces decentralized mechanisms of trust,transparency, and digital ownership. Current research addressing the intersection of thesedomains is fragmented and predominantly conceptual, leaving critical gaps in trust mechanisms,governance structures, operational models, and legal frameworks.To address these gaps, this study proposes the conceptual AIChain Framework: a unifiedplatform that integrates GenAI-powered content generation, automated tokenization, trustengines, and decentralized marketplaces. This architecture demonstrates cross-sectoral potentialin creative industries, FinTech, and education by linking algorithmic creativity withdecentralized ownership. The contributions are threefold: (1) at the theoretical level, the studysynthesizes the Resource-Based View (RBV), the Dynamic Capabilities View (DCV), the digitaltrust framework, and the information interaction model to establish a foundation for analyzingGenAI–Web3 convergence; (2) at the practical level, it introduces an operational architecture fornext-generation platform development; and (3) at the policy and governance level, it highlightsthe need for transparent, auditable, and participatory models to prevent technological oligopolies.By bridging theoretical insights with practical implications, this research provides a roadmap forfuture scholarship and industry practice, including pilot implementations of the AIChainframework, the design of hybrid governance models, and the assessment of ethical andenvironmental implications surrounding GenAI–Web3 convergence.
Forecasting cryptocurrency prices is hindered by extreme volatility and a methodological dilemma between information-scarce univariate models and noise-prone full-multivariate models. This paper investigates a partial-multivariate approach to balance this trade-off, hypothesizing that a strategic subset of features offers superior predictive power. We apply the Partial-Multivariate Transformer (PMformer) to forecast daily returns for BTCUSDT and ETHUSDT, benchmarking it against eleven classical and deep learning models. Our empirical results yield two primary contributions. First, we demonstrate that the partial-multivariate strategy achieves significant statistical accuracy, effectively balancing informative signals with noise. Second, we experiment and discuss an observable disconnect between this statistical performance and practical trading utility; lower prediction error did not consistently translate to higher financial returns in simulations. This finding challenges the reliance on traditional error metrics and highlights the need to develop evaluation criteria more aligned with real-world financial objectives.
This paper introduces a hybrid framework for portfolio optimization that fuses Long Short-Term Memory (LSTM) forecasting with a Proximal Policy Optimization (PPO) reinforcement learning strategy. The proposed system leverages the predictive power of deep recurrent networks to capture temporal dependencies, while the PPO agent adaptively refines portfolio allocations in continuous action spaces, allowing the system to anticipate trends while adjusting dynamically to market shifts. Using multi-asset datasets covering U.S. and Indonesian equities, U.S. Treasuries, and major cryptocurrencies from January 2018 to December 2024, the model is evaluated against several baselines, including equal-weight, index-style, and single-model variants (LSTM-only and PPO-only). The framework's performance is benchmarked against equal-weighted, index-based, and single-model approaches (LSTM-only and PPO-only) using annualized return, volatility, Sharpe ratio, and maximum drawdown metrics, each adjusted for transaction costs. The results indicate that the hybrid architecture delivers higher returns and stronger resilience under non-stationary market regimes, suggesting its promise as a robust, AI-driven framework for dynamic portfolio optimization.