Publications
An Improved DenseNet Model for Pediatric Pneumonia Detection Using Chest X-Ray Images
Mar 11, 2026Journal INTERNATIONAL JOURNAL ON INFORMATICS VISUALIZATION
DOI http://dx.doi.org/10.62527/joiv.10.1.5114
Issue No 1
Volume Vol 10
Pneumonia is a respiratory disorder that involves inflammation of the air sacs of the lungs and is normally diagnosed through imaging of the chest using X-rays. This research proposes a deep learning-based classification system to classify chest X-ray images and identify pneumonia. The methodological framework adopted is the Cross-Industry Standard Process for Data Mining (CRISP-DM), in which experiments are run in a Jupyter Notebook using five-fold cross-validation. The data is made up of anterior-posterior chest Xray photographs of children aged between one and five years of age at Guangzhou Women and Children Medical Center. A number of convolutional neural network models are assessed and compared with the proposed Improved DenseNet (ImDenseNet), including DenseNet, VGG16, and InceptionNet. According to experimental findings, ImDenseNet achieves 96.15% accuracy, 92.86% precision, and 92.94% recall, which are significantly better than those of the base models. The results show that the proposed architectural improvements improve feature discrimination and classification performance, and ImDenseNet is a credible solution for detecting pneumonia using chest X-ray images. Future research could focus on expanding ImDenseNet into a multi-class classifier to differentiate between bacterial and viral pneumonia. Pruning and quantization techniques can also be used to optimize the model for lightweight deployment on the edge or in clinical devices. Also, by incorporating explainable artificial intelligence (XAI) algorithms, clinical interpretability and confidence may be improved. Pneumonia is a respiratory disorder that involves inflammation of the air sacs of the lungs and is normally diagnosed through imaging of the chest using X-rays. This research proposes a deep learning-based classification system to classify chest X-ray images and identify pneumonia. The methodological framework adopted is the Cross-Industry Standard Process for Data Mining (CRISP-DM), in which experiments are run in a Jupyter Notebook using five-fold cross-validation. The data is made up of anterior-posterior chest X-ray photographs of children aged between one and five years of age at Guangzhou Women and Children Medical Center. A number of convolutional neural network models are assessed and compared with the proposed Improved DenseNet (ImDenseNet), including DenseNet, VGG16, and InceptionNet. According to experimental findings, ImDenseNet achieves 96.15% accuracy, 92.86% precision, and 92.94% recall, which are significantly better than those of the base models. The results show that the proposed architectural improvements improve feature discrimination and classification performance, and ImDenseNet is a credible solution for detecting pneumonia using chest X-ray images. Future research could focus on expanding ImDenseNet into a multi-class classifier to differentiate between bacterial and viral pneumonia. Pruning and quantization techniques can also be used to optimize the model for lightweight deployment on the edge or in clinical devices. Also, by incorporating explainable artificial intelligence (XAI) algorithms, clinical interpretability and confidence may be improved.
Federated Learning Architecture for Cross-Domain Threat Detection in Distributed IoT Networks with Dynamic Adversarial Adaptation
Mar 11, 2026Journal International Journal on Advanced Science, Engineering and Information Technology (IJASEIT)
DOI https://doi.org/10.18517/ijaseit.16.1.15443
Issue No. 1
Volume Vol. 16
The fast growth of the Internet of Things (IoT) infrastructure has provided tremendous amounts of networks of interconnected gadgets, each with a potential point of entry for advanced cyber threats. Traditional centralized security models are not able to handle the scale, diversity, and dynamics of the existing IoT settings, resulting in performance bottlenecks and privacy violations. As a way of overcoming these limitations, this paper introduces a Federated Learning (FL)-based Intrusion Detection Systems (IDS) to identify threats within distributed IoT networks in a secure and privacy-friendly way. The proposed system integrates cross-domain learning and adaptive adversarial systems, where IoT devices can cooperate and define new threats without sharing sensitive raw data. Experiments were conducted on real-world IoT sensor data, such as temperature and humidity, motion, gas, acceleration, light intensity, and GPS to test the applicability of IoT sensors to smart homes, industrial automation, and environmental monitoring. The system offered substantial enhancements in flexibility and communication effectiveness compared to centralized models by simulating attacks, including false data injection, signal interference, and digital tampering. The devices learn and exchange encrypted model updates independently, while keeping bandwidth to a minimum without affecting detection accuracy. The results show that the FL-based IDS can enhance the cybersecurity of large-scale IoT systems while ensuring data privacy and scalability across heterogeneous domains, making it a promising approach for implementing next-generation distributed cybersecurity systems. The fast growth of the Internet of Things (IoT) infrastructure has provided tremendous amounts of networks of interconnected gadgets, each with a potential point of entry for advanced cyber threats. Traditional centralized security models are not able to handle the scale, diversity, and dynamics of the existing IoT settings, resulting in performance bottlenecks and privacy violations. As a way of overcoming these limitations, this paper introduces a Federated Learning (FL)-based Intrusion Detection Systems (IDS) to identify threats within distributed IoT networks in a secure and privacy-friendly way. The proposed system integrates cross-domain learning and adaptive adversarial systems, where IoT devices can cooperate and define new threats without sharing sensitive raw data. Experiments were conducted on real-world IoT sensor data, such as temperature and humidity, motion, gas, acceleration, light intensity, and GPS to test the applicability of IoT sensors to smart homes, industrial automation, and environmental monitoring. The system offered substantial enhancements in flexibility and communication effectiveness compared to centralized models by simulating attacks, including false data injection, signal interference, and digital tampering. The devices learn and exchange encrypted model updates independently, while keeping bandwidth to a minimum without affecting detection accuracy. The results show that the FL-based IDS can enhance the cybersecurity of large-scale IoT systems while ensuring data privacy and scalability across heterogeneous domains, making it a promising approach for implementing next-generation distributed cybersecurity systems.
An energy-aware clustering protocol based grid for wsn
Jan 12, 2018Journal International Journal on Informatics Visualization
DOI 10.1088/1742-6596/1019/1/012007
Issue 4
Volume 2