International American University

Abstract Selecting an appropriate supportive response for a mental health question post is a key step toward scalable mental health question answering, yet it remains difficult due to semantic mismatch, noisy informal language, and the need to align supportive intent beyond surface lexical overlap. We propose MF-GAT, a novel multi view graph attention matching framework that constructs a Concept Interaction Graph (CIG) to explicitly encode post response concept alignments and their interaction structure. MF-GAT learns three complementary evidence streams, including local interaction features, multi view fusion features, and global context features, and applies view specific graph attention to propagate and reweight informative relational signals over the CIG. A gated fusion module then adaptively integrates the view representations into a unified matching vector for prediction. We evaluate MF-GAT both as a pair classification model and as a retrieval ranking model for selecting the best support from a candidate pool, reporting Accuracy and F1 together with standard ranking metrics including MRR and NDCG. On the MHQA benchmark, MF-GAT achieves 0.95 Accuracy and 0.85 F1, outperforming BERT (0.89, 0.67), CIG-GCN (0.92, 0.79), ARC-II (0.85, 0.60), and MatchPyramid (0.82, 0.59). These results show that novel multi view interaction graph modeling with attention based propagation improves both supportive response classification and practical retrieval quality for mental health support selection.

Climate variability and extreme rainfall events pose significant challenges to infrastructure development in Bangladesh, where standard statistical models often fail to account for nonlinear anomalies and low-frequency extremes. This study employs two machine learning approaches, Prophet and Long Short-Term Memory (LSTM) networks, to predict long-term yearly rainfall using historical data from 1980 to 2024 across two climatically sensitive locations. Prophet, an additive decomposition model, and LSTM, a recurrent neural network with memory-based learning, were benchmarked using Root Mean Squared Error (RMSE) and the Coefficient of Determination (R2). Results demonstrate that LSTM consistently outperformed Prophet in both Rajshahi and Ishwardi, obtaining lower RMSE values (102.4 mm and 118.7 mm, respectively) and higher R2 scores (0.88 and 0.85). While Prophet generated smoother forecasts, it underfitted severe years, whereas LSTM correctly captured interannual volatility and identified a greater number of high-risk years above the 90th percentile threshold. The estimates were transformed into practical adaptation techniques, including elevated foundation design, decentralised rainwater collection, and the use of water-resistant materials, thereby integrating predictive analytics into civil engineering applications. This integration provides a framework for climate-resilient infrastructure planning aligned with national adaptation plans. Future studies should expand to include multivariate forecasting that incorporates exogenous climate drivers, such as ENSO, soil moisture, and vegetation indices, as well as examine the scalable integration of GIS and BIM platforms for real-time urban resilience design.

The digital transformation of the U.S. economy has intensified the need for robust financial forecasting and proactive cybersecurity solutions. This study integrates advanced machine learning techniques—Long Short-Term Memory (LSTM), Prophet, and Random Forest Regressor—for financial time-series forecasting across twelve leading U.S. companies. In parallel, a cybersecurity framework was developed using the CICIDS2017 dataset and trained with a Random Forest Classifier to detect cyber intrusions. The forecasting models demonstrated high accuracy, with Random Forest achieving the best performance across most companies, while LSTM showed strong consistency in sequential trend learning. The cybersecurity model achieved an accuracy of 99.89% and an F1-score of 99.72%, effectively identifying attack patterns. Finally, a conceptual “AI Risk Intelligence Decision System” is proposed, fusing financial risk indicators with cyber threat intelligence to support early decision-making in digital financial operations. The proposed framework enhances predictive accuracy and cyber resilience, demonstrating strong potential for integration into real-world financial and cyber risk monitoring systems.

Breast cancer etiology is multifactorial with African American women experiencing a significant health disparity in clinical presentation and outcomes. The selenium-containing protein SELENOF has been implicated in breast carcinogenesis by cell culture and animal studies. SELENOF translation is highly regulated in part by the RNA helicase eIF4a3, which binds to the key regulatory regions in the SELENOF mRNA and suppress its translation. In addition, SELENOP, the primary selenium transporter, plays a critical role in selenium delivery to tissues and may influence selenoprotein synthesis. This study aimed to examine the levels of SELENOF and eIF4a3, along with SELENOF and SELENOP genotypes, in breast cancer tissues from African American and Caucasian women To study their roles in breast cancer outcome and racial disparity, human tissues were assessed by multiplex immunofluorescence staining with antibodies directed against SELENOF and eIF4a3 and DNA from these tissues were genotyped for previously studied variations in SELENOF and the selenium transporter protein SELENOP Elevated levels of both SELENOF and eIF4a3 were observed in breast cancer tissues. SELENOF expression and genotype varied by HER2 status, while SELENOP genotypes were associated with breast cancer and showed age-related differences. SELENOF and eIF4a3 were also higher in tissues derived from African American women, who also exhibited higher frequency of a SELENOP polymorphism in the non-coding region of the gene These findings suggest that SELENOF, eIF4a3, and SELENOP may contribute to breast cancer progression and racial disparities in outcomes. Their differential expression and genetic variation highlight potential molecular mechanisms underlying these disparities and may inform future therapeutic or diagnostic strategies.

The Kingdom of Saudi Arabia (KSA) boasts a significant potential for solar power generation due to its abundant solar energy resources. This solar energy potential can reduce KSA’s reliance on oil, provide a reliable power supply, create new employment opportunities, and contribute to a cleaner environment. Hence, for the proper utilization and deployment of solar energy applications, a long-term statistical understanding of the availability of sunshine intensity is critical, both in temporal and spatial domains. Accurate forecasting of solar resources is essential for effective integration of photovoltaics into the national grid. This study presents a hybrid framework integrating Multilayer Perceptron Artificial Neural Network (MPANN) and RMSprop Optimization (RMSO) for estimating the hour-ahead Global Horizontal Irradiance (GHI) for six locations in Saudi Arabia: Jubail, Al Jouf, Hagl, Riyadh, Jedda, and Najran. RMSO is employed to determine the optimal selection of MPANN hyperparameters to maximize prediction accuracy. The models analyse the stepwise reduction of input parameters in GHI prediction models to assess performance under diversified input feature sets. WindGHI-4 includes four input parameters: wind speed, air temperature, hour of the day, and previous-hour GHI. TempGHI-3 excludes wind speed, using three inputs, while LiteGHI-2 further simplifies the inputs to only air temperature and the hour of the day. The findings indicated the hybrid MPANN-RMSO model based WindGHI-4 features outperformed TempGHI-3 and LiteGHI-2 for GHI forecasting. The simulated results revealed that the deterministic coefficient is estimated as 0.993-0.983 for WindGHI-4, compared to 0.990-0.980 for TempGHI-3, and 0.972 and 0.905 for LiteGHI-2, respectively, for GHI prediction at six different zones in Saudi Arabia. The findings enable high-accuracy, hour-ahead GHI forecasting across diverse Saudi climatic conditions, providing robust decision support for solar power integration, improved grid stability, and more reliable operational planning.