Ahmed Moursi

Dr. Ahmed Moursi

School of Computer Science, University of Galway, Galway, Ireland
+1
orcid

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
29 July 2021 in Complex & Intelligent Systems
Verified
Journal Article
Open access
PDF file
Ahmed MoursiAhmed Moursi
Ahmed Moursi
NENawal El-Fishawy
Nawal El-Fishawy
Soufiene DjahelSoufiene Djahel
Soufiene Djahel
MSmarwa ahmed shouman
marwa ahmed shouman
Air pollution is a major issue resulting from the excessive use of conventional energy sources in developing countries and worldwide. Particulate Matter less than 2.5 µm in diameter (PM2.5) is the most dangerous air pollutant invading the human respiratory system and causing lung and heart diseases. Therefore, innovative air pollution forecasting methods and systems are required to reduce such risk. To that end, this paper proposes an Internet of Things (IoT) enabled system for monitoring and predicting PM2.5 concentration on both edge devices and the cloud. This system employs a hybrid prediction architecture using several Machine Learning (ML) algorithms hosted by Nonlinear AutoRegression with eXogenous input (NARX). It uses the past 24 h of PM2.5, cumulated wind speed and cumulated rain hours to predict the next hour of PM2.5. This system was tested on a PC to evaluate cloud prediction and a Raspberry Pi to evaluate edge devices’ prediction. Such a system is essential, responding quickly to air pollution in remote areas with low bandwidth or no internet connection. The performance of our system was assessed using Root Mean Square Error (RMSE), Normalized Root Mean Square Error (NRMSE), coefficient of determination (R2), Index of Agreement (IA), and duration in seconds. The obtained results highlighted that NARX/LSTM achieved the highest R2 and IA and the least RMSE and NRMSE, outperforming other previously proposed deep learning hybrid algorithms. In contrast, NARX/XGBRF achieved the best balance between accuracy and speed on the Raspberry Pi.

11 June 2022 in Sensors
Verified
Journal Article
Open access
PDF file
Ahmed MoursiAhmed Moursi
Ahmed Moursi
NENawal El-Fishawy
Nawal El-Fishawy
Soufiene DjahelSoufiene Djahel
Soufiene Djahel
MSmarwa ahmed shouman
marwa ahmed shouman
In a world where humanity’s interests come first, the environment is flooded with pollutants produced by humans’ urgent need for expansion. Air pollution and climate change are side effects of humans’ inconsiderate intervention. Particulate matter of 2.5 µm diameter (PM2.5) infiltrates lungs and hearts, causing many respiratory system diseases. Innovation in air pollution prediction is a must to protect the environment and its habitants, including those of humans. For that purpose, an enhanced method for PM2.5 prediction within the next hour is introduced in this research work using nonlinear autoregression with exogenous input (NARX) model hosting a convolutional neural network (CNN) followed by long short-term memory (LSTM) neural networks. The proposed enhancement was evaluated by several metrics such as index of agreement (IA) and normalized root mean square error (NRMSE). The results indicated that the CNN–LSTM/NARX hybrid model has the lowest NRMSE and the best IA, surpassing the state-of-the-art proposed hybrid deep-learning algorithms.

7 December 2019 in Menoufia Journal of Electronic Engineering Research
Verified
Journal Article
Closed access
Ahmed MoursiAhmed Moursi
Ahmed Moursi
MSMarwa A. Shouman
Marwa A. Shouman
EHEzz El-Din Hemdan
Ezz El-Din Hemdan
NENawal El-Fishawy
Nawal El-Fishawy
Air pollution is a phenomenon harmful to both human being existence as well as the ecological system. It is caused by the excess of some substances above a particular concentration in the atmosphere. Atmospheric particulate matter (APM) – or PM for short – threatens life because of its tiny size – diameter is up to 10 micrometers. Their danger comes from their ability to penetrate deeper inside the human respiratory system. (PM2.5) particulates are less than 2.5 micrometers and are more hazardous when compared to (PM10) coarse particles–10 micrometers in size. Hence, environmental agencies and governments seek to explore new methods to predict future air pollution. These endeavors mainly focus on mitigating environmental pollution and predicting pollutants concentrations to take enough precautions for citizens protection. This paper presents various machine learning algorithms that predict PM2.5 concentration for the next hour. These algorithms are Support Vector Regression (SVR), Long Short-Term Memory (LSTM), Random Forest, and Extra Trees. Their performance is measured in Root Mean Square Error (RMSE), coefficient of determination R2, and duration in seconds. Extra Trees shows the least RMSE and the highest coefficient of determination R2.

28 August 2024 in Neural Computing and Applications
Verified
Journal Article
Open access
PDF file
AFAlaa Fkirin
Alaa Fkirin
Ahmed MoursiAhmed Moursi
Ahmed Moursi
GAGamal Attiya
Gamal Attiya
Ayman EL-SAYEDAyman EL-SAYED
Ayman EL-SAYED
MSmarwa ahmed shouman
marwa ahmed shouman
Recent advancements in deep neural networks (DNNs) have made them indispensable for numerous commercial applications. These include healthcare systems and self-driving cars. Training DNN models typically demands substantial time, vast datasets and high computational costs. However, these valuable models face significant risks. Attackers can steal and sell pre-trained DNN models for profit. Unauthorised sharing of these models poses a serious threat. Once sold, they can be easily copied and redistributed. Therefore, a well-built pre-trained DNN model is a valuable asset that requires protection. This paper introduces a robust hybrid two-level protection system for safeguarding the ownership of pre-trained DNN models. The first-level employs zero-bit watermarking. The second-level incorporates an adversarial attack as a watermark by using a perturbation technique to embed the watermark. The robustness of the proposed system is evaluated against seven types of attacks. These are Fast Gradient Method Attack, Auto Projected Gradient Descent Attack, Auto Conjugate Gradient Attack, Basic Iterative Method Attack, Momentum Iterative Method Attack, Square Attack and Auto Attack. The proposed two-level protection system withstands all seven attack types. It maintains accuracy and surpasses current state-of-the-art methods.

1 July 2016 in Menoufia Journal of Electronic Engineering Research
Verified
Journal Article
Closed access
Ahmed MoursiAhmed Moursi
Ahmed Moursi
NENawal A. El-Fishawy
Nawal A. El-Fishawy
SYSherin M. Youssef
Sherin M. Youssef
Social Information Retrieval (SIR) is a relatively new domain which uses social information to enhance information retrieval processes. To find more interesting search results, social behavior can indicate how much these results are interesting. Social interaction over Web 2.0 are used here to enhance ranking of web results in response to a query. A dataset from Open Directory Project (ODP) is used here to show the improvement of ranking. We propose the usage of normalization and social services weights to give better performance. The proposed framework gets data from various types of social info (social bookmarking, social news, social network, discovery engines). Data is parsed into fields and significant values are used in the ranking process. Precision and Mean Average Precision (MAP) are used to evaluate results. Simulation results show better ranking with the proposed model.

24 October 2015
Verified
Conference Paper
Closed access
Ahmed MoursiAhmed Moursi
Ahmed Moursi
NEN. El-Fishawy
N. El-Fishawy
SYS. M. Youssef
S. M. Youssef
Information retrieval is a critical issue for everyone online. Searching for information happens all the time. To better retrieve our needs from the web, it’s logical to leverage data generated by users via Web 2.0 (web of interaction). Users’ actions - such as tagging a page, liking, sharing or other kinds of interaction - are used in this paper to better enhance information retrieval. The work presented in this paper used a dataset from Open Directory Project (ODP) to test the promising use of social services for better ranking of web results. We propose a framework to process social info and use it in the indexing and ranking processes. The proposed framework retrieves data from different kinds of social info (social bookmarking, social news, social network, discovery engines). Raw data is parsed into fields and significant values are either inserted into the index or used in the ranking process. Mean Average Precision (MAP) is used to evaluate results. Results are promising for using social tagging services such as Delicious. An initial assessment of the usage of other social services -for instance Facebook and so on- reveals that their data are not good enough to improve ranking of webpages.