Atif Ali Wagan

The escalating global prevalence of diabetes renders effective screening for Diabetic Retinopathy (DR) indispensable to prevent irreversible vision loss. Although deep learning models, particularly Convolutional Neural Networks (CNNs), attain diagnostic accuracy comparable to that of human experts, their black-box nature erodes clinical trust. To harmonize accuracy with interpretability, this paper proposes a novel CNN architecture that reformulates DR grading as a regression task. By substituting traditional dense layers with a Global Average Pooling (GAP) layer, our approach substantially reduces model complexity and training time while enabling the generation of Regression Activation Maps (RAMs). These RAMs deliver visual explanations by precisely highlighting the pathological regions that underpin the model's predictions. Evaluated on the Kaggle Diabetic Retinopathy Detection dataset, our model—through the replacement of dense layers with Global Average Pooling—markedly lowers model complexity while delivering diagnostic performance on par with baseline models employing fully-connected layers. The resulting system provides a simpler, more precise, and transparent alternative for automated medical screening, directly associating predictions with clinically relevant features.

Breast cancer is one of the most prevalent and life-threatening diseases among women worldwide. Accurate diagnosis from histopathological biopsy samples is essential, yet manual examination is time-consuming and subject to inter-observer variability, particularly given the shortage of trained pathologists alongside the increasing number of cases. Deep learning, especially Convolutional Neural Networks (CNNs), has emerged as a powerful tool for classifying medical images by automatically extracting discriminative features from raw data. In this study, we investigate the use of the publicly available Breast Cancer Histopathological (BreakHis) image database, which contains benign and malignant samples across multiple magnifications. Our proposed approach extracts isolated image patches, applies CNN-based feature learning and integrates multi-resolution information to improve classification performance. To enhance generalization under limited data, we adopt transfer learning with optimized fine-tuning. Experiments implemented in MATLAB demonstrate that our CNN-based framework achieves higher accuracy than traditional machine learning approaches relying on handcrafted texture descriptors. These findings highlight the potential of CNNs, combined with patch-based multi-resolution analysis, to support pathologists in reliable and efficient breast cancer diagnosis.

Effective collaboration among software developers relies heavily on their ability to communicate efficiently, with emotions playing a pivotal role in this process. Emotions are widely used in human decision‐making, making automated tools for emotion classification within developer communication channels essential. These tools can enhance productivity and collaboration by increasing awareness of fellow developers' emotions. Previous approaches, such as HOMER, RAKEL, and EmoTxt, have been proposed to classify emotions in Stack Overflow and Jira datasets at a finer granularity. However, these tools face performance challenges. To address these limitations, we aim to enhance multilabeled emotion classification performance by leveraging TextCNN, word embeddings, and hyper‐parameter optimization. We validate the performance of this method by comparing it with the best previous methods for emotion classification in software engineering text. This approach achieves an F1‐Micro score of 84.6001% on the Jira dataset and 76.9366% on the Stack Overflow dataset, showing an improvement of 3.5001% and 8.6366%, respectively. This advancement underscores the potential of this method in improving emotion classification performance, thereby fostering better collaboration and productivity among software developers.

The DNS firewall plays an important role in network security. It is based on a list of known malicious domain names, and, based on these lists, the firewall blocks communication with these domain names. However, DNS firewalls can only block known malicious domain names, excluding communication with unknown malicious domain names. Prior research has found that machine learning techniques are effective for detecting unknown malicious domain names. However, those methods have limited capabilities to learn from both textual and numerical data. To solve this issue, we present a novel unified learning approach that uses both numerical and textual features of the domain name to classify whether a domain name pair is malicious or not. The experiments were conducted on a benchmark domain names dataset consisting of 90,000 domain names. The experimental results show that the proposed approach performs significantly better than the six comparative methods in terms of accuracy, precision, recall, and F1-Score.