Analysis of Diabetes Clinical Data Based on Recurrent Neural Networks

Computational Intelligence and Neuroscience, Journal Year: 2022, Volume and Issue: 2022, P. 1 - 7

Published: June 27, 2022

At present, diabetes is one of the most important chronic noncommunicable diseases, that have threatened human health. By 2020, number diabetic patients worldwide has reached 425 million. This amazing attracted great attention various countries. With progress computing technology, many mathematical models and intelligent algorithms been applied in different fields health care. 822 subjects were selected this paper. They divided into 389 423 nondiabetic patients. Each included 41 indicators. Too indicator variables would increase computational effort there could be a strong correlation data redundancy between data. Therefore, sample features first dimensionally reduced to generate seven new space, retaining up 99.9% valid information from original A diagnostic classification model for clinical based on recurrent neural networks constructed, particle swarm optimization (PSO) was introduced optimise network's hyperparameters achieve effective diagnosis diabetes.

Language: Английский

---

Secured Entry Control: An IoT and Deep Learning based Mask Detection System for COVID-19 Suspected Patient Screening and Access Regulation

Published: April 25, 2024

The new coronavirus SARS-CoV-2, which triggered the COVID-19 pandemic, has had an unparalleled effect on economies, cultures, and world health. In response to critical need for strict screening systems in public areas, this study presents a creative Secured Entry Control system. Detecting controlling possible carriers attempting enter country is made by system, makes use of deep learning algorithms IoT technology. mask detection algorithm, MobileNetV2 model, exceptional validation accuracy $98.96 \%$. model's reliability supported performance evaluations using ROC curves, confusion matrix analysis, AUC value \%$, close optimal score $100 Because well-suited low-processing devices, it easy deploy Raspberry Pi, helps create affordable Furthermore, spotting increased body temperatures, contactless temperature sensor improves system's ability identify carriers. functioning system confirmed working prototype that presented Experimental Results section. main goal research autonomous selectively allows access people who are less likely transmit areas. To achieve overall objective reducing spread spaces, highlights successful integration identification IoT.

Language: Английский

---

Efficient Face Mask Detection Using Hybrid Deep Learning Algorithms

Journal of Al-Qadisiyah for Computer Science and Mathematics, Journal Year: 2024, Volume and Issue: 16(4)

Published: Dec. 30, 2024

The coronavirus COVID-19 pandemic has caused a global health crisis. According to the World Health Assembly, one of best preventative measures is wear face mask while out outdoors (WHO). This work presents hybrid model for identification that combines deep and traditional machine learning. I have trained proposed system, which consists convolutional neural networks (ConNN), support vector machines (SVM), random forests (RF), in three stages, first stage, used ConNN, second same ConNN with SVM method, third RF. paper suggests different kinds masked recognition datasets: Incorrectly Masked Face Dataset (IMFD), Correctly (CMFD), combination MaskedFace-Net, worldwide detection system. Two objectives are presented realistic i) identify individuals whose faces covered or not covered, ii) masks put on properly improperly (for example, at airport entrances among crowds). suggested made up two parts. part designed feature extraction using networks. In contrast, section classify RF methods. achieved 99.92%. 99.94%. 98.79%. Moreover,The system been tested real world scenarios can recognize any image selected by Google high accuracy. we comparison results aim evaluate model.

Language: Английский

---

A Survey Of Machine Learning Techniques For Detecting Anomaly In Internet Of Things (IoT)

Journal of Independent Studies and Research - Computing, Journal Year: 2023, Volume and Issue: 21(1)

Published: June 7, 2023

In recent years, there has been a lot of focus on anomaly detection. Technological advancements, such as the Internet Things (IoT), are rapidly being acknowledged critical means for data streams that create massive amounts in real time from variety applications. Analyzing this gathered to detect abnormal occurrences helps decrease functional hazards and avoid unnoticed errors cause programme delay. Methods evaluating specific anomalous behaviorsin IoT stream sources have established developed current literature. Unfortunately, very few thorough researches include all elements acquisition. As result, article seeks address void by presenting comprehensive picture numerous cutting-edge solutions fundamental concerns essential issues data. The type, types anomalies,the learning method, datasets, evaluation criteria described. Lastly, necessitate further investigation future approaches highlighted.

Language: Английский

---

Convolutional Neural Network-based Architecture for Detecting Face Mask in Crowded Areas

Published: July 2, 2023

After the invasion of Covid-19 virus, governments started containing spread virus by forcing people to wear face masks in public places. Therefore, automatic mask detection has become very important limit spread. Unfortunately, existing methods present limited performance accurately detecting crowded areas due significant number faces per scene. In order tackle this challenge, we propose a two-stage neural network-based architecture that can detect environments. Several simulations have been conducted investigate efficiency proposed and results show high accuracy reach up 96.5%.

Language: Английский

---