Publications

[1] N. I. S. Mohammad, “Data Scheduling Algorithm for Scalable and Efficient IoT Sensing in Cloud Computing,” arXiv:2508.04334, 2025. [Online]. Available: https://arxiv.org/abs/2508.04334

[2]. N. I. S. Mohammad, “Deep Spectral Epipolar Representations for Dense Light Field Reconstruction,” arXiv:2508.08900, 2025. [Online]. Available: https://arxiv.org/abs/2508.08900

[3]. N. I. S. Mohammad, “Hierarchical Spatial Algorithms for High-Resolution Image Quantization and Feature Extraction,” arXiv:2510.08449, 2025. [Online]. Available: https://arxiv.org/abs/2510.08449

[4]. N. I. S. Mohammad, “A Multimodal XAI Framework for Trustworthy CNNs and Bias Detection in Deep Representation Learning,” arXiv:2510.12957, 2025. [Online]. Available: https://arxiv.org/abs/2510.12957

[5]. N. I. S. Mohammad, “Extended LSTM: Adaptive Feature Gating for Toxic Comment Classification,” arXiv:2510.17018, 2025. [Online]. Available: https://arxiv.org/abs/2510.17018

[6]. Optimized Sampling and Feature Selection Framework for Robust PCOS Diagnosis Using Extreme Gradient Boosting

Abstract: Polycystic Ovary Syndrome (PCOS) is a leading cause of infertility in women, with long-term implications that extend beyond reproductive health, including heightened risks of metabolic and cardiovascular disorders. Accurate and early detection is therefore critical for effective clinical management. This study presents a hybrid diagnostic framework that leverages the power of Extreme Gradient Boosting (XGBoost) for robust PCOS classification. To mitigate class imbalance and noisy samples within the dataset, we applied a resampling strategy combining Synthetic Minority Oversampling Technique (SMOTE) with Edited Nearest Neighbor (ENN) filtering. Furthermore, feature selection was conducted using statistical correlation analysis, Chi-Square, and ANOVA tests, identifying 23 highly discriminative clinical and metabolic indicators. Experimental evaluation on a publicly available benchmark dataset demonstrated that our optimized XGBoost model consistently outperformed baseline classifiers. The proposed method achieved an average cross-validation accuracy of 96.03% and a recall of 98% for non-PCOS cases, establishing new performance benchmarks compared to previously reported approaches. These findings underscore the potential of combining advanced feature engineering with ensemble-based gradient boosting to support reliable, data-driven PCOS diagnosis and enhance decision-making in clinical practice. Code

[1]. N. I. S. Mohammad, S. Tuna, B.U. Töreyin, “Optimized Sampling and Feature Selection Framework for Robust PCOS Diagnosis Using Extreme Gradient Boosting,” Manuscript in preparation for submission to a Q1 journal in Computational Health Informatics, 2025.

[7]. Probabilistic Ensemble Learning for Accurate Breast Cancer Diagnosis Using Machine Learning Algorithms

Abstract: Breast cancer is one of the most prevalent invasive cancers and a leading cause of cancer-related mortality among women. Early detection is critical for improving survival rates, which has motivated extensive research into automated diagnostic systems. This study presents a hybrid machine learning framework that combines predictions from three base classifiers, Logistic Regression, Support Vector Machine, and K-Nearest Neighbors, to enhance their performance through stacked Gradient Boosting. This approach achieved a 10-fold cross-validation accuracy of 98.4%, recall of 100%, and precision of 97.3%. To address class imbalance, SMOTE (Synthetic Minority Oversampling Technique) was applied, and robust scaling was used for normalization and outlier handling. Our framework integrates domain-specific knowledge at every stage, from data preprocessing and feature selection to model training and classification, ensuring robust and interpretable results. Comparative evaluation demonstrates that the proposed method outperforms several recent state-of-the-art approaches, highlighting its potential for reliable, data-driven breast cancer diagnosis. Code

[2]. N. I. S. Mohammad, S. Tuna, B.U. Toreyin, “Probabilistic Ensemble Learning for Accurate Breast Cancer Diagnosis Using Machine Learning Algorithms,” Manuscript in preparation for submission to a Q1 journal in Computational Health Informatics, 2025.