Mohammed Awad
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Mohammed Awad has a total of 35 co-authors. Here are the top co-authors by number of shared publications:
- Hector Pomares
Department of Computer Engineering, Automation and Robotics, CITIC-UGR, University of Granada, 18071 Granada, Spain
7 shared publications - Luis Javier Herrera
University of Granada
6 shared publications - Ignacio Rojas
Department of Computer Engineering, Automation and Robotics, CITIC-UGR, University of Granada, 18071 Granada, Spain
5 shared publications - Jesús González
University of Granada
5 shared publications A. Herrera
2 shared publications- Alaa Sheta
Southern Connecticut State University
2 shared publications
Latest Publications
- Open Access
- 2 April 2024
- Published by CSIR-National Institute of Science Communication and Policy Research (NIScPR) in Journal of Scientific & Industrial Research
- Vol. 83 (4) , 432-444
A Hybrid Model of Neural Networks and Genetic Algorithms for Prediction of the Stock Market Price: Case Study of Palestine
Accurate stock market predictions are critical to investor protection and economic growth. This study is the first of its kind to anticipate Palestinian stock market values using artificial intelligence models. In this paper, an improved hybrid model is given that combines multilayer perceptron neural networks with genetic algorithms to predict the state of the Palestinian stock market using the Al-Quds Index as the major indicator (MLPNNs-GAs). Furthermore, the stock values of the three largest Palestinian companies will be forecast using their stock market data. The rationale for merging artificial neural networks (ANNs) and genetic algorithms (GAs) stem from the fact that stock price data bear highly volatile and nonlinear features. The undiscovered patterns of relationships in the input and output data can be explored by artificial neural networks. The weights for the NNs are optimized using genetic algorithms (GAs), which determine the optimal weights based on performance and best-predicted minimal mean square error (MSE) value. Recurrent neural networks with Levenberg-Marquardt (RNNs-LM) and MLPNNs-LM, two more classic models of various neural network techniques, were used to compare the prediction performance of the proposed model in terms of mean square error. The experimental results show that, with MSEs of 0.0011 for the Al-Quds Index, 0.0021 for the Bank of Palestine, 0.001 for Palte, and 0.0006 for Padico, the recommended hybrid model MLPNNs-GAs outperforms other models in terms of closing price predictions. It has been shown that the MLPNNs-GAs model may give stock market investors reliable and accurate tools for making forecasts; as a result, MLPNNs-GAs is advised as an effective model for the prediction of nonlinear financial time series data.
- 2 April 2024
- Published by CSIR-National Institute of Science Communication and Policy Research (NIScPR) in Journal of Scientific & Industrial Research
- Vol. 83 (4) , 432-444
Open AccessA Hybrid Model of Neural Networks and Genetic Algorithms for Prediction of the Stock Market Price: Case Study of Palestine
Accurate stock market predictions are critical to investor protection and economic growth. This study is the first of its kind to anticipate Palestinian stock market values using artificial intelligence models. In this paper, an improved hybrid model is given that combines multilayer perceptron neural networks with genetic algorithms to predict the state of the Palestinian stock market using the Al-Quds Index as the major indicator (MLPNNs-GAs). Furthermore, the stock values of the three largest Palestinian companies will be forecast using their stock market data. The rationale for merging artificial neural networks (ANNs) and genetic algorithms (GAs) stem from the fact that stock price data bear highly volatile and nonlinear features. The undiscovered patterns of relationships in the input and output data can be explored by artificial neural networks. The weights for the NNs are optimized using genetic algorithms (GAs), which determine the optimal weights based on performance and best-predicted minimal mean square error (MSE) value. Recurrent neural networks with Levenberg-Marquardt (RNNs-LM) and MLPNNs-LM, two more classic models of various neural network techniques, were used to compare the prediction performance of the proposed model in terms of mean square error. The experimental results show that, with MSEs of 0.0011 for the Al-Quds Index, 0.0021 for the Bank of Palestine, 0.001 for Palte, and 0.0006 for Padico, the recommended hybrid model MLPNNs-GAs outperforms other models in terms of closing price predictions. It has been shown that the MLPNNs-GAs model may give stock market investors reliable and accurate tools for making forecasts; as a result, MLPNNs-GAs is advised as an effective model for the prediction of nonlinear financial time series data.
- Open Access
- 30 March 2024
- Published by Elsevier in Measurement: Energy
- Vol. 2
Time series forecasting of electricity consumption using hybrid model of recurrent neural networks and genetic algorithms
- 30 March 2024
- Published by Elsevier in Measurement: Energy
- Vol. 2
Open AccessTime series forecasting of electricity consumption using hybrid model of recurrent neural networks and genetic algorithms
- Open Access
FedCSD: A Federated Learning Based Approach for Code-Smell Detection
This paper proposes a Federated Learning Code Smell Detection (FedCSD) approach that allows organizations to collaboratively train federated ML models while preserving their data privacy. These assertions have been supported by three experiments that have significantly leveraged three manually validated datasets aimed at detecting and examining different code smell scenarios. In experiment 1, which was concerned with a centralized training experiment, dataset two achieved the lowest accuracy (92.30%) with fewer smells, while datasets one and three achieved the highest accuracy with a slight difference (98.90% and 99.5%, respectively). This was followed by experiment 2, which was concerned with cross-evaluation, where each ML model was trained using one dataset, which was then evaluated over the other two datasets. Results from this experiment show a significant drop in the model's accuracy (lowest accuracy: 63.80\%) where fewer smells exist in the training dataset, which has a noticeable reflection (technical debt) on the model's performance. Finally, the last and third experiments evaluate our approach by splitting the dataset into 10 companies. The ML model was trained on the company's site, then all model-updated weights were transferred to the server. Ultimately, an accuracy of 98.34% was achieved by the global model that has been trained using 10 companies for 100 training rounds. The results reveal a slight difference in the global model's accuracy compared to the highest accuracy of the centralized model, which can be ignored in favour of the global model's comprehensive knowledge, lower training cost, preservation of data privacy, and avoidance of the technical debt problem.
Open AccessFedCSD: A Federated Learning Based Approach for Code-Smell Detection
This paper proposes a Federated Learning Code Smell Detection (FedCSD) approach that allows organizations to collaboratively train federated ML models while preserving their data privacy. These assertions have been supported by three experiments that have significantly leveraged three manually validated datasets aimed at detecting and examining different code smell scenarios. In experiment 1, which was concerned with a centralized training experiment, dataset two achieved the lowest accuracy (92.30%) with fewer smells, while datasets one and three achieved the highest accuracy with a slight difference (98.90% and 99.5%, respectively). This was followed by experiment 2, which was concerned with cross-evaluation, where each ML model was trained using one dataset, which was then evaluated over the other two datasets. Results from this experiment show a significant drop in the model's accuracy (lowest accuracy: 63.80\%) where fewer smells exist in the training dataset, which has a noticeable reflection (technical debt) on the model's performance. Finally, the last and third experiments evaluate our approach by splitting the dataset into 10 companies. The ML model was trained on the company's site, then all model-updated weights were transferred to the server. Ultimately, an accuracy of 98.34% was achieved by the global model that has been trained using 10 companies for 100 training rounds. The results reveal a slight difference in the global model's accuracy compared to the highest accuracy of the centralized model, which can be ignored in favour of the global model's comprehensive knowledge, lower training cost, preservation of data privacy, and avoidance of the technical debt problem.
- Open Access
- 25 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 44888-44904
FedCSD: A Federated Learning Based Approach for Code-Smell Detection
Software quality is critical, as low quality, or "Code smell," increases technical debt and maintenance costs. There is a timely need for a collaborative model that detects and manages code smells by learning from diverse and distributed data sources while respecting privacy and providing a scalable solution for continuously integrating new patterns and practices in code quality management. However, the current literature is still missing such capabilities. This paper addresses the previous challenges by proposing a Federated Learning Code Smell Detection (FedCSD) approach, specifically targeting "God Class," to enable organizations to train distributed ML models while safeguarding data privacy collaboratively. We conduct experiments using manually validated datasets to detect and analyze code smell scenarios to validate our approach. Experiment 1, a centralized training experiment, revealed varying accuracies across datasets, with dataset two achieving the lowest accuracy (92.30%) and datasets one and three achieving the highest (98.90% and 99.5%, respectively). Experiment 2, focusing on cross-evaluation, showed a significant drop in accuracy (lowest: 63.80%) when fewer smells were present in the training dataset, reflecting technical debt. Experiment 3 involved splitting the dataset across 10 companies, resulting in a global model accuracy of 98.34%, comparable to the centralized model’s highest accuracy. The application of federated ML techniques demonstrates promising performance improvements in code-smell detection, benefiting both software developers and researchers.
- 25 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 44888-44904
Open AccessFedCSD: A Federated Learning Based Approach for Code-Smell Detection
Software quality is critical, as low quality, or "Code smell," increases technical debt and maintenance costs. There is a timely need for a collaborative model that detects and manages code smells by learning from diverse and distributed data sources while respecting privacy and providing a scalable solution for continuously integrating new patterns and practices in code quality management. However, the current literature is still missing such capabilities. This paper addresses the previous challenges by proposing a Federated Learning Code Smell Detection (FedCSD) approach, specifically targeting "God Class," to enable organizations to train distributed ML models while safeguarding data privacy collaboratively. We conduct experiments using manually validated datasets to detect and analyze code smell scenarios to validate our approach. Experiment 1, a centralized training experiment, revealed varying accuracies across datasets, with dataset two achieving the lowest accuracy (92.30%) and datasets one and three achieving the highest (98.90% and 99.5%, respectively). Experiment 2, focusing on cross-evaluation, showed a significant drop in accuracy (lowest: 63.80%) when fewer smells were present in the training dataset, reflecting technical debt. Experiment 3 involved splitting the dataset across 10 companies, resulting in a global model accuracy of 98.34%, comparable to the centralized model’s highest accuracy. The application of federated ML techniques demonstrates promising performance improvements in code-smell detection, benefiting both software developers and researchers.
- 26 September 2023
- Published by Elsevier in Expert Systems with Applications
- Vol. 238
Enhanced variants of crow search algorithm boosted with cooperative based island model for global optimization
- 26 September 2023
- Published by Elsevier in Expert Systems with Applications
- Vol. 238
Enhanced variants of crow search algorithm boosted with cooperative based island model for global optimization
- 23 September 2023
- Published by Elsevier in Biomedical Signal Processing and Control
- Vol. 87
Hybrid deep learning and metaheuristic model based stroke diagnosis system using electroencephalogram (EEG)
- 23 September 2023
- Published by Elsevier in Biomedical Signal Processing and Control
- Vol. 87
Hybrid deep learning and metaheuristic model based stroke diagnosis system using electroencephalogram (EEG)
- Open Access
- 28 May 2022
- Published by Pensoft Publishers in JUCS - Journal of Universal Computer Science
- Vol. 28 (5) , 499-539
An Enhanced Evolutionary Based Feature Selection Approach Using Grey Wolf Optimizer for the Classification of High-dimensional Biological Data
Feature selection (FS) is a pre-processing step that aims to eliminate the redundant and less-informative features to enhance the performance of data mining techniques. It is also considered as one of the key success factors for classification problems in high-dimensional datasets. This paper proposes an efficient wrapper feature selection method based on Grey Wolf Optimizer (GWO). GWO is a recent metaheuristic algorithm that has been widely employed to solve diverse optimization problems. However, GWO mainly follows the search directions toward the leading wolves, making it prone to fall into local optima, especially when dealing with high-dimensional problems, which is the case when dealing with many biological datasets. An enhanced variation of GWO called EGWO, which adapts two enhancements, is introduced to overcome this specific shortcoming. In the first place, the transition parameter concept is incorporated to move GWO from the exploration phase to the exploitation phase. Several adaptive non-linear decreasing formulas are introduced to control the transition parameters. In the second place, a random-based search strategy is exploited to empower diversity during the search process. Two binarization schemes using S-shaped and V-shaped transfer functions are incorporated to map the continuous search space into a binary one for FS. The efficiency of the proposed EGWO is validated on ten high-dimensional low-samples biological data. Our experiments show the promising performance of EGWO compared to the original GWO approach and other state-of-the-art techniques in terms of dimensionality reduction and the enhancement of classification performance.
- 28 May 2022
- Published by Pensoft Publishers in JUCS - Journal of Universal Computer Science
- Vol. 28 (5) , 499-539
Open AccessAn Enhanced Evolutionary Based Feature Selection Approach Using Grey Wolf Optimizer for the Classification of High-dimensional Biological Data
Feature selection (FS) is a pre-processing step that aims to eliminate the redundant and less-informative features to enhance the performance of data mining techniques. It is also considered as one of the key success factors for classification problems in high-dimensional datasets. This paper proposes an efficient wrapper feature selection method based on Grey Wolf Optimizer (GWO). GWO is a recent metaheuristic algorithm that has been widely employed to solve diverse optimization problems. However, GWO mainly follows the search directions toward the leading wolves, making it prone to fall into local optima, especially when dealing with high-dimensional problems, which is the case when dealing with many biological datasets. An enhanced variation of GWO called EGWO, which adapts two enhancements, is introduced to overcome this specific shortcoming. In the first place, the transition parameter concept is incorporated to move GWO from the exploration phase to the exploitation phase. Several adaptive non-linear decreasing formulas are introduced to control the transition parameters. In the second place, a random-based search strategy is exploited to empower diversity during the search process. Two binarization schemes using S-shaped and V-shaped transfer functions are incorporated to map the continuous search space into a binary one for FS. The efficiency of the proposed EGWO is validated on ten high-dimensional low-samples biological data. Our experiments show the promising performance of EGWO compared to the original GWO approach and other state-of-the-art techniques in terms of dimensionality reduction and the enhancement of classification performance.
- Open Access
- 8 September 2019
- Published by MECS Publisher in International Journal of Intelligent Systems and Applications
- Vol. 11 (9) , 40-55
Prediction of Water Demand Using Artificial Neural Networks Models and Statistical Model
The prediction of future water demand will help water distribution companies and government to plan the distribution process of water, which impacts on sustainable development planning. In this paper, we use a linear and nonlinear models to predict water demand, for this purpose, we will use different types of Artificial Neural Networks (ANNs) with different learning approaches to predict the water demand, compared with a known type of statistical methods. The dataset depends on sets of collected data (extracted from municipalities databases) during a specific period of time and hence we proposing a nonlinear model for predicting the monthly water demand and finally provide the more accurate prediction model compared with other linear and nonlinear methods. The applied models capable of making an accurate prediction for water demand in the future for the Jenin city at the north of Palestine. This prediction is made with a time horizon month, depending on the extracted data, this data will be used to feed the neural network model to implement mechanisms and system that can be employed to predicts a short-term for water demands. Two applied models of artificial neural networks are used; Multilayer Perceptron NNs (MLPNNs) and Radial Basis Function NNs (RBFNNs) with different learning and optimization algorithms Levenberg Marquardt (LM) and Genetic Algorithms (GAs), and one type of linear statistical method called Autoregressive integrated moving average ARIMA are applied to the water demand data collected from Jenin city to predict the water demand in the future. The execution results appear that the MLPNNs-LM type is outperformed the RBFNN-GAs and ARIMA models in the prediction the water demand values.
- 8 September 2019
- Published by MECS Publisher in International Journal of Intelligent Systems and Applications
- Vol. 11 (9) , 40-55
Open AccessPrediction of Water Demand Using Artificial Neural Networks Models and Statistical Model
The prediction of future water demand will help water distribution companies and government to plan the distribution process of water, which impacts on sustainable development planning. In this paper, we use a linear and nonlinear models to predict water demand, for this purpose, we will use different types of Artificial Neural Networks (ANNs) with different learning approaches to predict the water demand, compared with a known type of statistical methods. The dataset depends on sets of collected data (extracted from municipalities databases) during a specific period of time and hence we proposing a nonlinear model for predicting the monthly water demand and finally provide the more accurate prediction model compared with other linear and nonlinear methods. The applied models capable of making an accurate prediction for water demand in the future for the Jenin city at the north of Palestine. This prediction is made with a time horizon month, depending on the extracted data, this data will be used to feed the neural network model to implement mechanisms and system that can be employed to predicts a short-term for water demands. Two applied models of artificial neural networks are used; Multilayer Perceptron NNs (MLPNNs) and Radial Basis Function NNs (RBFNNs) with different learning and optimization algorithms Levenberg Marquardt (LM) and Genetic Algorithms (GAs), and one type of linear statistical method called Autoregressive integrated moving average ARIMA are applied to the water demand data collected from Jenin city to predict the water demand in the future. The execution results appear that the MLPNNs-LM type is outperformed the RBFNN-GAs and ARIMA models in the prediction the water demand values.