Forecasting Stock Prices: A Machine Learning-Based Approach for Predictive Analytics Through a Case Study

Stock price prediction has always been a challenging task, requiring careful observation of trends and dynamics of the market because of the volatile and complex nature of financial markets. Various factors affect market behavior all the time. Even some unquantifiable factors like 25 Oct 2025 (Published Online) emotions of the masses, social and political dynamics, etc., also play a great role. So perfect Machine Learning, Deep Learning, behaviors into consideration is crucial for better prediction of the ups and downs of prices. SMA, EMA, RSI, MACD, Bollinger Various machine learning and deep learning models have been proposed to tackle the challenges Bands, RFE, Random Forest by capturing and interpreting complex patterns and relationships in historical price data. Regressor, Multivariate Analysis, Technical features are important for understanding market trends and thus improving the LSTM. accuracy of stock price predictions. In this paper, we calculate key technical indicators such as Simple Moving Average (SMA), Exponential Moving Average (EMA), Relative Strength Index (RSI), Moving Average Convergence Divergence (MACD), Bollinger Bands, and others. We then focus on selecting the most relevant indicators by employing feature selection methods from these to enhance the extraction of meaningful features reflecting underlying market behavior and increase the probability of more precise prediction. Here, Recursive Feature Elimination (RFE) and Random Forest Regressor-based importance ranking methods have been applied for the feature selection task. To get a better forecast of market price, it is important to capture long- term dependencies and patterns over time. Long Short-Term Memory (LSTM) networks are well- suited for modeling and predicting sequential data like stock prices. By leveraging an LSTM model and taking the selected features, we do a multivariate analysis to forecast stock price based on historical data, identifying the trends fairly accurately with some lags here and there.

Technology-Assisted Parent Training Programs for Autism Management

By Rayhan Khan

The developmental condition known as autism spectrum disorder (ASD) is defined by recurring behavioural patterns and challenges with social communication. Taking care of a kid with impairments presents parents with a lot of emotional and practical 19 Aug 2024 (Published Online) obstacles that might affect their family's arrangements. This article examines the integration and efficacy of technology-based parenting interventions for addressing ASD, focusing on how these programs are developed, which technologies are used, and Technology assisted program, Parent how they affect parent-child relations and success rates. The phenomenology design, a teacher collaboration. qualitative research approach, was used to analyse the experiences of primary school students with disabilities in virtual education activities after the global pandemic 2020. The design allowed for a comprehensive understanding of students' perspectives and solutions. Face-to-face training techniques are effective but cannot reach all families due to transport, money, and time issues. Distance-based training and technology- assisted training solutions provide a solution by disseminating high-quality, evidence- based training to a broader audience. The results show that ADEPT and the PLAY Project are examples of potential supports involving the application of digital tools to provide parents with essential training content to create proper home conditions for further child development. Evaluating the success of these initiatives is crucial to assessing their impact and potentially modifying them. Scientific methods like randomised controlled trials or longitudinal studies provide insights into the efficacy of technology-supported training. At the same time, measurable quantities like parent- child interaction or behavioural changes prove its effectiveness.

Blockchain-Based Banking Infrastructure for Securing Financial Transactions and Reducing Operational Costs in the U.S.

Blockchain technology is rapidly reshaping the financial landscape in the United States, offering improved security, transparency, and operational efficiency across banking, trade finance, and regulatory compliance. This study explores how U.S. financial institutions are integrating 25 October, 2025 (Published Online) blockchain to enhance performance, drawing on a range of case studies from both public and Blockchain Technology, Financial a 42% reduction in fraudulent transactions, a 58% decrease in trade finance settlement times, and Inclusion, Cybersecurity, Trade a 49% boost in compliance efficiency. In addition, blockchain is playing a critical role in Finance, Artificial Intelligence, protecting against cyber threats, with blockchain-secured institutions reporting a 47% drop in Regulatory Compliance cyberattacks and a 31% improvement in fraud detection through the use of AI-integrated blockchain systems. Mobile blockchain applications have also increased banking accessibility, particularly in underserved areas, supporting broader financial inclusion efforts. Furthermore, the convergence of blockchain with emerging technologies such as artificial intelligence (AI), the Internet of Things (IoT), and cloud computing has enabled real-time transaction monitoring, secure data sharing, and more robust trade verification processes. Despite ongoing challenges related to regulatory clarity and system integration, blockchain is emerging as a foundational technology in the U.S. financial system, with strong potential to drive innovation, strengthen cybersecurity, and create a more inclusive and efficient financial ecosystem.

Digital Transformation in Leadership Management: Opportunities and Challenges in the COVID-19 Scenario

Digital transformation is a crucial process in the development of public and private organizations, with various conceptions. This paper aims to understand the perception, cognitive development, positive aspects, achievements, urgency, and challenges faced 15 Aug 2024 (Published Online) by civil servants and leaders in the digital transformation process. The research method is based on available documents and authors' views. The findings will contribute to the theoretical basis and direction for leaders, highlighting the importance of leadership thinking innovation in driving successful digital transformation across countries, especially emerging ones. The research demonstrates the challenges of digital transformation and the need for effective leadership.

Green Finance and Its Impact on Sustainable Investment Strategies in the US

This research aims to analyze green finance's applicability in forming sustainable investment policies in the USA. This research fills a literature gap to uncover the long-run equilibrium co- integration between FDI inflows, CO2 emissions, renewable energy, and renewable electricity. 25 Oct 2025 (Published Online) Using VECM and Johansen co-integration tests, this paper discusses the long-run relationship of Green Finance, Foreign Direct gathered from the World Bank. The analysis reveals that the two variables are co-integrated over Investment (FDI), Sustainable the long run, though there is a short-run time-varying co-integration relationship. For instance, Investment, Renewable Energy, and the co-integration test results present a trace statistic of 72.77, and its p-value is 0.0001, which CO₂ Emissions justifies the existence of co-integration, which is a long-term equilibrium. The IRF analysis also shows that renewable energy consumption positively affects FDI, and levels off at 0.28 after 4 periods, whereas CO2 emissions have a negative long-run effect on FDI with a coefficient of - 4.9153. Based on these findings, applying green finance policies for renewable energy import can encourage foreign investments in the short run. However, the cost involved in shifting to renewable energy sources may lead to a restricted number of long-term investments. This motivated the study to recommend a search for more information on such sector dynamics.

Fraud Transaction Detection using Machine Learning on Financial Datasets

Financial fraud poses a significant threat to the digital economy, with credit card fraud being a prevalent challenge. This study evaluates the performance of Logistic Regression (LR) and Extreme Gradient Boosting (XG Boost) models in detecting fraudulent transactions using 25 Oct 2025 (Published Online) financial datasets. The study uses practical data from 284,807 transactions, but only 492 are Fraud Detection, Machine Learning, Technique (SMOTE). Our findings show that XG Boost with Random Search selection is better XGBoost, Logistic Regression, and than Logistic Regression in all aspects. XG Boost yielded an accuracy of 99.96%, precision of Imbalanced Dataset (SMOTE) 95.11%, recall of 79.61%, and F1 score of 86.61%, while for Logistic Regression, the corresponding percentages were 99.92%, 88.1%, 60.5%, and 71.7%. The AUC statistic of 0.98 for XG Boost against 0.97 for LR classified the model as having better discriminant power. The results show that XG Boost is more suitable for real-time fraud detection. However, computational limitations and explainability issues should be considered. For future work, it is suggested that semi-supervised and supervised learning approaches be investigated and work with larger datasets to improve fraud detection in financial systems.