VFAST - Virtual Foundation for Advancement of Science and Technology (Pakistan)
Not a member yet
1255 research outputs found
Sort by
Financial News Sentiment Analysis Using NLP and Machine Learning for Asset Price Prediction: A Systematic Review
Full text linkForecasting market movements in stocks, gold, and crude oil requires a deep understanding of how financial news sentiment influences asset prices. Analyzing news sentiment is crucial for understanding market dynamics and forecasting price fluctuations. However, creating accurate financial news datasets, particularly in terms of proper labeling and sourcing, continues to be a significant challenge. This paper presents a comprehensive literature review on financial news sentiment analysis and its application in market trend prediction.By reviewing articles in reputable journals from 2018–2025, we consolidate key findings, including techniques for dataset creation, labeling, and sourcing, as well as the use of advanced methods such as Natural Language Processing (NLP) and deep learning models. This review contributes to the growing literature on sentiment analysis in the context of the relationship between stocks and commodities, especially gold, crude oil, and the role of global and market specific news sentiments in determining the assets prices. The study focuses on issues that concern researchers in this regard; it also compares the relative success of various prediction models and discusses the criteria for assessing their effectiveness.We propose solutions to current challenges and outline future research directions to improve sentiment analysis in financial markets
Double Hierarchy Linguistic Soft and Soft Covering Based Sets and Their Application
Full text linkRecent developments in mathematics have resulted in the creation of the soft set theory as a method for dealing with uncertainty. The classical soft sets, however, are not well suitable for dealing with double hierarchy linguistic term parameters. This study\u27s objective is to expand traditional soft sets to include double hierarchy linguistic soft sets and double hierarchy linguistic soft covering based sets, whereby the DHLTSs, soft sets, and soft covering-based sets merge. Next, we describe upper and lower approximation, positive, negative, and border regions for double hierarchy linguistic soft sets and double hierarchy linguistic soft covering based sets. Ultimately, the double hierarchy linguistic soft set is applied to a decision-making issue with the help of fuzzy soft sets, and the value is demonstrated with a numerical example. In this example, we chose the best suitable Renewable Energy Source (RES) alternative for installation at Abdul Wali Khan University Mardan (AWKUM), using the recently established MCDM approach VIKOR method
Pakistan’s Economic Diplomacy: Balancing Trade and Economic Security
Full text linkThis essay highlights Pakistan\u27s shift from traditional geopolitics to a geo-economic approach while providing a full analysis of the evolving dynamics of its economic diplomacy. At the core of this change are important initiatives like the Economic Outreach Initiative and the China-Pakistan Economic Corridor (CPEC), which aim to promote trade and tourism, boost foreign investment, and strengthen regional integration. A key advantage of Pakistan is its strategic location, which enables it to manage complex ties with neighboring nations like India and Afghanistan while also interacting with regional powers like China, the United States, Iran, and members of the Shanghai Cooperation Organization (SCO). The study explores the intricate relationship between economic diplomacy (ED) and economic security (ES) using a qualitative methodology that includes case studies, questionnaires, and documentary analysis. Using realism, constructivism and Economic Statecraft The paper demonstrates Pakistan\u27s efforts to balance commercial liberalization with national security needs. The results demonstrate Pakistan\u27s strength and ability to benefit from free trade agreements (FTAs) and regional cooperation, despite significant challenges such as political unrest, poor infrastructure, and non-tariff constraints. The article concludes with recommendations for changing policies, making targeted investments, and committing to sustainable development in order to secure Pakistan\u27s economic success in a rapidly evolving global marketplace
Understanding the Educational and Social Influences on Future Technopreneurs’ Engagement with the Gerontechnology Market
Full text linkThe research will explore the interplay of education and social factors in the culture of future technopreneurs when it comes to the perception and plans about the currently emerging gerontechnology market. However, with the number of aging population across the globe increasing, there is dire need to develop new solutions that are tailored towards the needs of older adults. Based on a mixed-methods research design, the research synthesis involves quantitative survey data of the study of students at universities and qualitative research based on interviews. The goal is to measure the awareness, motivation, and readiness of students to take part in gerontechnology entrepreneurship. It is revealed that entrepreneurial education, values of the society, and experience of demographic challenges and burdens play an important role in defining interests and attitudes of students to this branch. Many of the participants showed interest in joining efforts in socially significant innovations, especially with the provision of applicable knowledge and related real-world working experiences. This research reveals the importance of including the aging content into the teaching entrepreneurship and encourage intergenerational work. Through these actions, it is possible to raise a new breed of socially responsible technopreneurs out of educational institutions who are fully prepared to embrace the opportunities and challenges brought by an aging population. This strategy will not only accommodate the needs of the market but will also promote the development of sustainable solutions that will improve the quality of life of the older adult
Hybrid EfficientNet Models with Attention Mechanisms for Enhanced Bone Fracture Detection and Classification Using X-ray Images
Full text linkBone fractures are a major diagnostic problem that needs rapid and prompt diagnosis. This paper presents an improved Efficient Net-B03 model and a self-attention mechanism to capture the local and global features for detection and classification. Data were obtained on the Kaggle bone fracture dataset. The model improves focus on fine details of fractures produced with computational efficiency. The preprocessing methods were adopted to enhance the quality of X-ray images, and the data augmentation was applied, and transfer learning also enhanced performance. The Efficient Net-B03 and the Attention Mechanism were trained separately and subsequently combined to come up with the proposed Hybrid model. The proposed hybrid model and Efficient Net-B03 both reached an accuracy of 0.99. The classification reports show that the dataset was categorized into two classes: Class 0 (non-fracture) had a precision of 0.99, and a recall. With a 0.88, Class 1 (fracture) had a 0.95 precision, 1.00 is recall and F1-scores is 0.93 for Class 0 and 0.97 for Class 1, with support values of 372 and 925 instances, respectively. Overall, the proposed hybrid model achieved an accuracy range from 0.96 to 0.98 with cross-validation, and up to 0.99 accuracy without cross-validation. The proposed hybrid model will help in the radiology clinical setting
Protection and Authentication of DICOM Images with Iris Recognition
Full text linkProtection of medical images over the internet is a challenging task regarding its communication and usefulness. This work proposes a novel mechanism for the protection of DICOM images. The technique provides the facility to embed the iris (image) as a watermark to protect its integrity. After retrieval of the watermark, the mechanism is so robust that it recognizes the iris efficiently and correctly. Similarly, the system protects the Region of Interest (ROI) before the extraction of watermark as it is quite difficult to find volume, visualize or analyze the ROI. The technique not only protects the watermark but also gives the protection to ROI, which may be different in various medical images. Experimental results provide a deep view about the robustness of the technique in real time against both the intentional and unintentional attacks. The empirical results also show that after extraction of the watermark, the system has the capability to accurately find the volume, visualize and analyze the ROI
Next Word Prediction for Urdu using Deep Learning Techniques
Full text linkA language model for next-word prediction is a probabilistic representation of a natural language that utilizes text corpora to generate word probabilities. These models play a crucial role in text generation, machine translation, and question-answering applications. The focus of this study is to develop an improved algorithm for next-word prediction in Urdu. The study implements deep learning models, including RNN, LSTM, and Bi-LSTM, on a subset of the Ur-Mono Urdu corpus containing 3,000 and 5,000 sentences. To prepare the data for experimentation, tokenization and stemming data cleaning techniques are applied. The study achieved an accuracy of 87% using the RNN model on the first 3,000 sentences of the Ur-Mono dataset and 84% accuracy using the RNN model on the first 5,000 sentences of the Ur-Mono dataset. In conclusion, it can be stated that when the corpus size is small, the RNN outperforms both the LSTM and BiLSTM. However, as the corpus size increases, the Bi-LSTM exhibits superior performance compared to both RNN and LSTM
Data-Driven Student Performance Analysis: A Machine Learning Approach
No full textThe utilization of machine learning (ML) methods has the potential to address the challenges posed by the rapid growth of student-related data, enabling better predictions of student performance and supporting informed managerial decisions. These techniques analyze data through advanced models and algorithms to forecast academic outcomes. This research focuses on identifying key factors that influence student performance using ML approaches. By leveraging statistical and classification algorithms, machine learning enhances the accuracy of predictions. The research explores relevant factors and applies in state-of-art models to achieve precise performance predictions. Various studies have employed ML techniques to predict student success, highlighting its broad applicability. This research proposes a framework for assessing students\u27 academic achievements. The dataset includes information such as demographic details, prior academic records, and family background. Data was sourced from students across multiple universities using online surveys, comprising 24 attributes adapted from prior research. The objective is to identify the critical attributes that significantly affect student performance. It also evaluates distinguish classification techniques to enhance prediction accuracy. Experimental findings reveal that the Support Vector Machine (SVM) outperforms other methods, achieving a maximum accuracy of 62.50%. This research proposed the effective prediction tools can be developed to improve educational outcomes effectively
Investigating the Role of LASSO in Feature Selection for Educational Data Mining (EDM) Applications
Full text linkWith the advent of digitalization, education-related activities have started generating massive amounts of data from various facets, such as student interaction, assessment, and learning management systems. Such vast amounts of data become suitable areas for Educational Data Mining (EDM) to reveal insights for actionable improvement in academic outcomes and personalized learning experiences. However, high dimensionality and the redundancy of the educational data also pose considerable threats to the accuracy, interpretability, and computational efficiency of modeling. Least Absolute Shrinkage and Selection Operator (LASSO) is one powerful technique for simultaneous regression and feature selection. By introducing sparsity, LASSO minimizes the absolute sum of regression coefficients, thereby forcing insignificant features to be reduced to zero automatically. This feature is handy in EDM, where relevant indicators such as attendance, quiz scores, or study patterns must be distinguished from noisy or redundant variables. This paper systematically investigates the application of LASSO in EDM by giving the mathematical background and geometric interpretation, along with practical usage recommendations. Also, LASSO performance has been checked on synthetic and real datasets, including the famous dataset UCI Student Performance. The findings prove that LASSO significantly enhances model interpretability, predictive accuracy, and a decline in complexity. In conclusion, limitations are discussed, as well as practical considerations and future directions for LASSO applications to next-generation educational analytics
Enhancing Model Robustness in Federated Learning: A Systematic Literature Review of Byzantine-Resilient Aggregation Methods
Full text linkThe demand for privacy-preserving machine learning has led to the rise of Federated Learning (FL), where multiple clients collaboratively train a model without sharing raw data. Despite its privacy benefits, FL is vulnerable to Byzantine failures, where malicious or faulty participants inject corrupted updates, threatening model integrity. To address this, a range of Byzantine-resilient aggregation techniques have been proposed, including statistical filters (e.g., Trimmed Mean, Krum), trust-based weighting, cryptographic protocols, and hybrid strategies. This paper presents a systematic literature review (SLR) of these defenses, evaluating their robustness, scalability, and suitability for real-world applications. Challenges such as non-IID data, adaptive attacks, and trade-offs between security and efficiency are critically examined. In addition, we explore emerging trends such as domain-specific defenses, energy-aware FL, quantum-resilient methods, and federated zero-knowledge proofs. A novel classification of hybrid approaches and a standardized benchmarking framework are proposed to guide future research. This review aims to support the development of resilient, efficient and scalable decentralized learning systems in adversarial environments