Electronic copies of Computing Science and Mathematics journal articles. http://hdl.handle.net/1893/216 2026-10-03T07:06:33Z 2026-10-03T07:06:33Z Hotoğlu, Esra Sen, Sevil Can, Burcu http://hdl.handle.net/1893/38346 2026-09-21T09:57:21Z 2026-12-01T00:00:00Z

Title: A Comprehensive Analysis of Adversarial Attacks against Spam Filters Author(s): Hotoğlu, Esra; Sen, Sevil; Can, Burcu Abstract: Deep learning has revolutionized email filtering, which is critical to protect users from cyber threats such as spam, malware, and phishing. However, the increasing sophistication of adversarial attacks poses a significant challenge to the effectiveness of these filters. This study investigates the impact of adversarial attacks on deep learning-based spam detection systems using real-world datasets. Six prominent deep learning models are evaluated on these datasets, analyzing attacks at the word, character sentence, and AI-generated paragraph-levels. Novel scoring functions, including spam weights and attention weights, are introduced to improve attack effectiveness. A key contribution of this study is the analysis of spam-weight-and attention-weight-based scoring functions, highlighting their role in improving the effectiveness and efficiency of adversarial attacks. This comprehensive analysis sheds light on the vulnerabilities of spam filters and contributes to efforts to improve their security against evolving adversarial threats. Experimental results show that word-and sentence-level attacks markedly increase false negatives, while character-level perturbations disrupt token representations with minimal semantic change. AI-generated paragraph-level attacks remain challenging even for transformer-based models. In addition, spam-weight-based scoring consistently enables more effective adversarial attacks than alternative scoring strategies with lower computational cost.

2026-12-01T00:00:00Z Hapca, Simona Yang, Qinbo Li, Sheyu McGurnaghan, Stuart J Blackbourn, Luke A K Pearson, Ewan R Colhoun, Helen M Bell, Samira http://hdl.handle.net/1893/38322 2026-09-18T00:26:39Z 2026-09-01T00:00:00Z

Title: The Impact of Non-AKI eGFR Variability on CKD Progression in Individuals With Type 2 Diabetes and Preserved Kidney Function Author(s): Hapca, Simona; Yang, Qinbo; Li, Sheyu; McGurnaghan, Stuart J; Blackbourn, Luke A K; Pearson, Ewan R; Colhoun, Helen M; Bell, Samira Abstract: Introduction Variability in estimated glomerular filtration rate (eGFR) has been associated with increased risks of mortality and chronic kidney disease (CKD) progression in people with type 2 diabetes mellitus (T2DM) and impaired kidney function. However, its significance in individuals with preserved kidney function remains unclear. Methods In this nationwide retrospective population-based study of individuals with T2DM, eGFR variability was calculated by fitting a linear regression model to longitudinal data to estimate both the individual eGFR slope over the 5-year period as well as the variability in model residuals provided by the SD of the model residuals using longitudinal serum creatinine (SCr) measurements obtained during the first 5 years after diagnosis. Cox proportional hazards models were then applied to assess the association between eGFR variability and progression to stage G3b CKD among participants with preserved kidney function. Results This study included 98,322 participants who had an eGFR > 60 ml/min per 1.73 m2 at diagnosis, remained alive with an eGFR > 60 ml/min per 1.73 m2 5 years after diagnosis, and were subsequently followed for a mean of 5.1 years. Greater eGFR variability was associated with an increased risk of progression to stage G3b CKD- hazard ratios (HRs) for the second, third, and fourth quartiles of variability versus the first quartile were 1.56 (95% confidence interval [CI]: 1.38-1.75), 1.85 (95% CI: 1.65-2.08), and 2.56 (95% CI: 2.29-2.86), respectively. This association persisted after adjustment for multiple variables-HR: 1.57; 95% CI: 1.40-1.77 for the fourth quartiles of variability versus the first quartile. Conclusion eGFR variability in the absence of acute kidney injury (AKI) is associated with CKD progression in individuals with T2DM and preserved kidney function.

2026-09-01T00:00:00Z Kareem, Muhammad Shoaib Amjad, Madiha Aslam, Saba Rasool, Abdur Jamil, Mutiullah Ali, Hazrat http://hdl.handle.net/1893/38315 2026-09-18T00:22:27Z 2026-08-01T00:00:00Z

Title: GCE: A Framework for Interpretable Nonlinear Hazard Modeling in Cardiac Sarcoma Survival Using SEER Data Author(s): Kareem, Muhammad Shoaib; Amjad, Madiha; Aslam, Saba; Rasool, Abdur; Jamil, Mutiullah; Ali, Hazrat Abstract: Cardiac sarcoma is a rare aggressive malignancy for which survival prediction is limited by small cohorts, censoring, nonlinear prognostic effects, and incomplete interpretability. We developed the GRU–CoxPH Ensemble (GCE), a weighted late-fusion survival framework combining a single-step GRU static-covariate learner with Cox proportional hazards (CoxPHs). Using SEER Research Plus data, 727 eligible patients were identified from 41 source variables, and 27 raw SEER predictors were retained after leakage exclusion and Cox-LASSO/Random Survival Forest screening. Outcome, follow-up, cause-of-death, identifier, and endpoint-derived fields were removed before preprocessing; imputation, encoding, scaling, feature selection, tuning, and ensemble-weight selection were performed within training folds only. In leakage-free 10-fold out-of-fold evaluation, the GCE achieved high cohort-specific internal discrimination that is likely optimistic and requires external validation (mean C-index =0.9198; IBS = 0.03641), whereas CoxPH showed lower discrimination (C-index = 0.8605) but better IBS calibration (0.03378). Thus, the GCE is better for discrimination, under internal validation, and is not calibration-superior. SHAP provided post hoc descriptive association summaries for the final ensemble risk score. These internal results support the GCE only as a research framework for cardiac sarcoma survival-risk modeling; external validation and recalibration are required before any patient-level clinical consideration.

2026-08-01T00:00:00Z Gu, Yuanlin Zhang, Mao Li, Baihua Meng, Qinggang http://hdl.handle.net/1893/38257 2026-08-27T00:03:39Z 2026-08-01T00:00:00Z

Title: Interpretable machine learning for shared feature identification and time series prediction Author(s): Gu, Yuanlin; Zhang, Mao; Li, Baihua; Meng, Qinggang Abstract: When building a predictive model with sub-datasets from diverse locations, scenarios, or participants, a single model may not capture the unique characteristics of each sub-dataset. However, creating individual models for each dataset can be time-consuming and may overlook shared features. In this article, a Common Structure Neural Network (CSNN) model is introduced to address these issues. The model includes a new feature selection layer that identifies critical shared factors influencing multiple outputs, allowing for a shared model structure and reduced training costs, while accurately representing the diversity within each sub-dataset. The effectiveness of the model is demonstrated through one simulation and two real-world case studies on air pollution and stock prices. The experiments show that the model improves prediction accuracy and efficiency compared to other methods. Additionally, it enhances interpretability by revealing correlations and interactions across different locations, offering valuable insights.

2026-08-01T00:00:00Z