Mohammed Ali Alshara
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mohammed Ali Alshara has a total of 56 co-authors. Here are the top co-authors by number of shared publications:
- Naveed Ahmad
College of Computer and Information Sciences, Prince Sultan University, Riyadh, Saudi Arabia
15 shared publications - Hassan Jalil Hadi
King Abdullah University of Science and Technology
7 shared publications - Yue Cao
Wuhan University
5 shared publications - Yasir Javed
Department of Software Engineering, Prince Sultan University, Riyadh, Saudi Arabia
4 shared publications - Arshad Iqbal
School of Computing Sciences, Pak-Austria Fachhochschule: Institute of Applied Sciences and Technology, Mang, Haripur 22620, Pakistan
3 shared publications - Shakir Khan
College of Computer and Information Sciences, Imam Mohammad Ibn Saud Islamic University (IMSIU), Riyadh, Saudi Arabia
3 shared publications
Latest Publications
- Open Access
- 12 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 73096-73113
Enhancing Detection of Anomaly-Based Intrusions in Smart Health Using Explainable AI
Anomaly-based threat detection is crucial for organizations. Anomalies are deviations from the norm and can sometimes be hard to identify in contextual cases. These anomalies may arise from an insider-based malicious intrusion. Anomalies are mostly unsupervised and sometimes require expert validation to confirm their occurrence. The black-box nature of unsupervised machine learning models offers little insight into the origin and details of the anomaly. Thus, there is a need for an explanation layer in anomaly detection systems to aid decision-makers in tracking anomalies. Anomaly-based threat detection is critical for smart health scenarios. In this research, we have utilized the electronic patient record to identify unsupervised anomalies.We have suggested adding an explainability layer that utilizes unsupervised learning. The results indicated that Isolation Forest outperformed the other methods on the clustering metrics, as validated by the explainability layer. Lime and Shap are used to visually explain the anomaly’s cause using feature contributions. User-friendly, explainable captions are generated for each anomaly instance, enabling non-technical users to easily understand the results from Lime and Shap. Three metrics, fidelity, sparsity, and stability, are used to validate the explanation layer, which explains each instance of an anomaly: fidelity requires R2 ≥ 0.70, sparsity favors a smaller number of features that provide user-friendly results, and stability is high. Finally, a Lime-Shap agreement is used to validate the results for each anomaly instance across all Explanations (all seeds). The proposed scheme adds trust to unsupervised learning results, thereby increasing user confidence in the anomaly detection algorithm’s outcomes and aiding decision-makers.
- 12 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 73096-73113
Open AccessEnhancing Detection of Anomaly-Based Intrusions in Smart Health Using Explainable AI
Anomaly-based threat detection is crucial for organizations. Anomalies are deviations from the norm and can sometimes be hard to identify in contextual cases. These anomalies may arise from an insider-based malicious intrusion. Anomalies are mostly unsupervised and sometimes require expert validation to confirm their occurrence. The black-box nature of unsupervised machine learning models offers little insight into the origin and details of the anomaly. Thus, there is a need for an explanation layer in anomaly detection systems to aid decision-makers in tracking anomalies. Anomaly-based threat detection is critical for smart health scenarios. In this research, we have utilized the electronic patient record to identify unsupervised anomalies.We have suggested adding an explainability layer that utilizes unsupervised learning. The results indicated that Isolation Forest outperformed the other methods on the clustering metrics, as validated by the explainability layer. Lime and Shap are used to visually explain the anomaly’s cause using feature contributions. User-friendly, explainable captions are generated for each anomaly instance, enabling non-technical users to easily understand the results from Lime and Shap. Three metrics, fidelity, sparsity, and stability, are used to validate the explanation layer, which explains each instance of an anomaly: fidelity requires R2 ≥ 0.70, sparsity favors a smaller number of features that provide user-friendly results, and stability is high. Finally, a Lime-Shap agreement is used to validate the results for each anomaly instance across all Explanations (all seeds). The proposed scheme adds trust to unsupervised learning results, thereby increasing user confidence in the anomaly detection algorithm’s outcomes and aiding decision-makers.
- Open Access
- 20 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Communications Society
- Vol. PP, 1
A Deterministic Game-Theoretic Handover Protocol for UAV-RSU Integrated Networks
The increasing demand for time-sensitive applications in Intelligent Transportation Systems (ITS) powered by Vehicular Ad Hoc Networks (VANETs) burdens the limited communication channels available to the Road Side Units (RSUs), particularly in their dense serving areas. One of the most widely used applications of Unmanned Aerial Vehicles (UAVs) is to leverage them as on-demand aerial base stations to offload RSUs and provide a superior air-to-ground Line-of-Sight (LoS) channel. However, managing resource allocation between RSUs and UAVs in a highly dynamic vehicular network requires a novel predictive strategy. Current systems fail to jointly optimize UAV trajectory, channel resource allocation between the RSU and the UAV, and handover decisions between them to meet ultra-low-latency constraints. This paper proposes the Deterministic Coordinated Trajectory and Resource (DCRT) Optimization Framework, comprising a Kalman Filter for predictive vehicle-state estimation and a Cooperative Game Theory approach for optimal handover decisions between the RSU and the UAV. The DCRT framework solves a joint optimization problem to minimize total service time while maximizing average area throughput. Simulation results demonstrate the effectiveness of DCRT, showing an average service time reduction of approximately 61% compared with uncoordinated random access (RAC) and 48% compared with probabilistic access schemes (PAS) under medium-to-high traffic loads, while achieving a maximum throughput improvement of approximately 11.7% over RAC and 34.6% over PAS in terms of successful vehicles per slot under medium traffic densities. These results validate the deterministic game-theoretic approach for reliable ultra-low-latency V2X communication.
- 20 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of the Communications Society
- Vol. PP, 1
Open AccessA Deterministic Game-Theoretic Handover Protocol for UAV-RSU Integrated Networks
The increasing demand for time-sensitive applications in Intelligent Transportation Systems (ITS) powered by Vehicular Ad Hoc Networks (VANETs) burdens the limited communication channels available to the Road Side Units (RSUs), particularly in their dense serving areas. One of the most widely used applications of Unmanned Aerial Vehicles (UAVs) is to leverage them as on-demand aerial base stations to offload RSUs and provide a superior air-to-ground Line-of-Sight (LoS) channel. However, managing resource allocation between RSUs and UAVs in a highly dynamic vehicular network requires a novel predictive strategy. Current systems fail to jointly optimize UAV trajectory, channel resource allocation between the RSU and the UAV, and handover decisions between them to meet ultra-low-latency constraints. This paper proposes the Deterministic Coordinated Trajectory and Resource (DCRT) Optimization Framework, comprising a Kalman Filter for predictive vehicle-state estimation and a Cooperative Game Theory approach for optimal handover decisions between the RSU and the UAV. The DCRT framework solves a joint optimization problem to minimize total service time while maximizing average area throughput. Simulation results demonstrate the effectiveness of DCRT, showing an average service time reduction of approximately 61% compared with uncoordinated random access (RAC) and 48% compared with probabilistic access schemes (PAS) under medium-to-high traffic loads, while achieving a maximum throughput improvement of approximately 11.7% over RAC and 34.6% over PAS in terms of successful vehicles per slot under medium traffic densities. These results validate the deterministic game-theoretic approach for reliable ultra-low-latency V2X communication.
- Open Access
- 18 December 2025
- Published by MDPI AG in Information
- Vol. 16 (12) , 1114
GenIIoT: Generative Models Aided Proactive Fault Management in Industrial Internet of Things
Detecting active failures is important for the Industrial Internet of Things (IIoT). The IIoT aims to connect devices and machinery across industries. The devices connect via the Internet and provide large amounts of data which, when processed, can generate information and even make automated decisions on the administration of industries. However, traditional active fault management techniques face significant challenges, including highly imbalanced datasets, a limited availability of failure data, and poor generalization to real-world conditions. These issues hinder the effectiveness of prompt and accurate fault detection in real IIoT environments. To overcome these challenges, this work proposes a data augmentation mechanism which integrates generative adversarial networks (GANs) and the synthetic minority oversampling technique (SMOTE). The integrated GAN-SMOTE method increases minority class data by generating failure patterns that closely resemble industrial conditions, increasing model robustness and mitigating data imbalances. Consequently, the dataset is well balanced and suitable for the robust training and validation of learning models. Then, the data are used to train and evaluate a variety of models, including deep learning architectures, such as convolutional neural networks (CNNs) and long short-term memory networks (LSTMs), and conventional machine learning models, such as support vector machines (SVMs), K-nearest neighbors (KNN), and decision trees. The proposed mechanism provides an end-to-end framework that is validated on both generated and real-world industrial datasets. In particular, the evaluation is performed using the AI4I, Secom and APS datasets, which enable comprehensive testing in different fault scenarios. The proposed scheme improves the usability of the model and supports its deployment in a real IIoT environment. The improved detection performance of the integrated GAN-SMOTE framework effectively addresses fault classification challenges. This newly proposed mechanism enhances the classification accuracy up to 0.99. The proposed GAN-SMOTE framework effectively overcomes the major limitations of traditional fault detection approaches and proposes a robust, scalable and practical solution for intelligent maintenance systems in the IIoT environment.
- 18 December 2025
- Published by MDPI AG in Information
- Vol. 16 (12) , 1114
Open AccessGenIIoT: Generative Models Aided Proactive Fault Management in Industrial Internet of Things
Detecting active failures is important for the Industrial Internet of Things (IIoT). The IIoT aims to connect devices and machinery across industries. The devices connect via the Internet and provide large amounts of data which, when processed, can generate information and even make automated decisions on the administration of industries. However, traditional active fault management techniques face significant challenges, including highly imbalanced datasets, a limited availability of failure data, and poor generalization to real-world conditions. These issues hinder the effectiveness of prompt and accurate fault detection in real IIoT environments. To overcome these challenges, this work proposes a data augmentation mechanism which integrates generative adversarial networks (GANs) and the synthetic minority oversampling technique (SMOTE). The integrated GAN-SMOTE method increases minority class data by generating failure patterns that closely resemble industrial conditions, increasing model robustness and mitigating data imbalances. Consequently, the dataset is well balanced and suitable for the robust training and validation of learning models. Then, the data are used to train and evaluate a variety of models, including deep learning architectures, such as convolutional neural networks (CNNs) and long short-term memory networks (LSTMs), and conventional machine learning models, such as support vector machines (SVMs), K-nearest neighbors (KNN), and decision trees. The proposed mechanism provides an end-to-end framework that is validated on both generated and real-world industrial datasets. In particular, the evaluation is performed using the AI4I, Secom and APS datasets, which enable comprehensive testing in different fault scenarios. The proposed scheme improves the usability of the model and supports its deployment in a real IIoT environment. The improved detection performance of the integrated GAN-SMOTE framework effectively addresses fault classification challenges. This newly proposed mechanism enhances the classification accuracy up to 0.99. The proposed GAN-SMOTE framework effectively overcomes the major limitations of traditional fault detection approaches and proposes a robust, scalable and practical solution for intelligent maintenance systems in the IIoT environment.
- 10 June 2025
- Published by MyJove Corporation
- No. 220,p. e67822
JoVE Video Dataset
Brain tumors represent a significant medical challenge, necessitating accurate and efficient detection methods for timely intervention. This work integrates several pretrained base models, such as VGG16, MobileNetV2, DenseNet121, InceptionV3, and ResNet50, to propose a novel method for brain tumor diagnosis. A streamlined and standardized technique has been proposed to accommodate various base models, ensuring consistency and ease of maintenance and facilitating model comparison. To amplify the variety of the training dataset and enhance model generalization, notable image augmentation methods like adjusting brightness and contrast are utilized. Further, an effective training pipeline utilizing data generators is designed to process large datasets efficiently while conserving computing power. The study conducted a thorough analysis using three different optimizers (Adam, Stochastic Gradient Descent, and Adamax) applied to each pretrained base model, with comprehensive adjustments of hyperparameters. Metrics like recall, accuracy, precision, F1-score, and confusion matrices are used to evaluate the model's performance, providing a comprehensive understanding of the model's behavior. A systematic comparison of each model's performance provided an in-depth examination of strengths and weaknesses, facilitating informed model selection and decision-making for brain tumor detection applications. MobileNetV2 achieved the highest overall performance with an accuracy of 96%, precision of 96%, recall of 94%, and an F1-score of 95% using the Adam optimizer. DenseNet121 and VGG16 also performed well, achieving accuracies of 95% and 94%, respectively. InceptionV3 demonstrated a slightly lower performance compared to the top-performing models, with an accuracy of 93%, precision of 93%, recall of 91%, and an F1-score of 92%. ResNet50 showed relatively lower performance with an accuracy of 77%, precision of 78%, recall of 76%, and an F1-score of 76%. These metrics demonstrate the robustness and efficacy of the proposed method for brain tumor detection.
- 10 June 2025
- Published by MyJove Corporation
- No. 220,p. e67822
JoVE Video Dataset
Brain tumors represent a significant medical challenge, necessitating accurate and efficient detection methods for timely intervention. This work integrates several pretrained base models, such as VGG16, MobileNetV2, DenseNet121, InceptionV3, and ResNet50, to propose a novel method for brain tumor diagnosis. A streamlined and standardized technique has been proposed to accommodate various base models, ensuring consistency and ease of maintenance and facilitating model comparison. To amplify the variety of the training dataset and enhance model generalization, notable image augmentation methods like adjusting brightness and contrast are utilized. Further, an effective training pipeline utilizing data generators is designed to process large datasets efficiently while conserving computing power. The study conducted a thorough analysis using three different optimizers (Adam, Stochastic Gradient Descent, and Adamax) applied to each pretrained base model, with comprehensive adjustments of hyperparameters. Metrics like recall, accuracy, precision, F1-score, and confusion matrices are used to evaluate the model's performance, providing a comprehensive understanding of the model's behavior. A systematic comparison of each model's performance provided an in-depth examination of strengths and weaknesses, facilitating informed model selection and decision-making for brain tumor detection applications. MobileNetV2 achieved the highest overall performance with an accuracy of 96%, precision of 96%, recall of 94%, and an F1-score of 95% using the Adam optimizer. DenseNet121 and VGG16 also performed well, achieving accuracies of 95% and 94%, respectively. InceptionV3 demonstrated a slightly lower performance compared to the top-performing models, with an accuracy of 93%, precision of 93%, recall of 91%, and an F1-score of 92%. ResNet50 showed relatively lower performance with an accuracy of 77%, precision of 78%, recall of 76%, and an F1-score of 76%. These metrics demonstrate the robustness and efficacy of the proposed method for brain tumor detection.
- 10 June 2025
- Published by MyJove Corporation in Journal of Visualized Experiments
- No. 220,p. e67822
Augmented Multimodal Fusion for Optimized Brain Tumor Detection: Evaluation and Comparative Analysis
Brain tumors represent a significant medical challenge, necessitating accurate and efficient detection methods for timely intervention. This work integrates several pretrained base models, such as VGG16, MobileNetV2, DenseNet121, InceptionV3, and ResNet50, to propose a novel method for brain tumor diagnosis. A streamlined and standardized technique has been proposed to accommodate various base models, ensuring consistency and ease of maintenance and facilitating model comparison. To amplify the variety of the training dataset and enhance model generalization, notable image augmentation methods like adjusting brightness and contrast are utilized. Further, an effective training pipeline utilizing data generators is designed to process large datasets efficiently while conserving computing power. The study conducted a thorough analysis using three different optimizers (Adam, Stochastic Gradient Descent, and Adamax) applied to each pretrained base model, with comprehensive adjustments of hyperparameters. Metrics like recall, accuracy, precision, F1-score, and confusion matrices are used to evaluate the model's performance, providing a comprehensive understanding of the model's behavior. A systematic comparison of each model's performance provided an in-depth examination of strengths and weaknesses, facilitating informed model selection and decision-making for brain tumor detection applications. MobileNetV2 achieved the highest overall performance with an accuracy of 96%, precision of 96%, recall of 94%, and an F1-score of 95% using the Adam optimizer. DenseNet121 and VGG16 also performed well, achieving accuracies of 95% and 94%, respectively. InceptionV3 demonstrated a slightly lower performance compared to the top-performing models, with an accuracy of 93%, precision of 93%, recall of 91%, and an F1-score of 92%. ResNet50 showed relatively lower performance with an accuracy of 77%, precision of 78%, recall of 76%, and an F1-score of 76%. These metrics demonstrate the robustness and efficacy of the proposed method for brain tumor detection.
- 10 June 2025
- Published by MyJove Corporation in Journal of Visualized Experiments
- No. 220,p. e67822
Augmented Multimodal Fusion for Optimized Brain Tumor Detection: Evaluation and Comparative Analysis
Brain tumors represent a significant medical challenge, necessitating accurate and efficient detection methods for timely intervention. This work integrates several pretrained base models, such as VGG16, MobileNetV2, DenseNet121, InceptionV3, and ResNet50, to propose a novel method for brain tumor diagnosis. A streamlined and standardized technique has been proposed to accommodate various base models, ensuring consistency and ease of maintenance and facilitating model comparison. To amplify the variety of the training dataset and enhance model generalization, notable image augmentation methods like adjusting brightness and contrast are utilized. Further, an effective training pipeline utilizing data generators is designed to process large datasets efficiently while conserving computing power. The study conducted a thorough analysis using three different optimizers (Adam, Stochastic Gradient Descent, and Adamax) applied to each pretrained base model, with comprehensive adjustments of hyperparameters. Metrics like recall, accuracy, precision, F1-score, and confusion matrices are used to evaluate the model's performance, providing a comprehensive understanding of the model's behavior. A systematic comparison of each model's performance provided an in-depth examination of strengths and weaknesses, facilitating informed model selection and decision-making for brain tumor detection applications. MobileNetV2 achieved the highest overall performance with an accuracy of 96%, precision of 96%, recall of 94%, and an F1-score of 95% using the Adam optimizer. DenseNet121 and VGG16 also performed well, achieving accuracies of 95% and 94%, respectively. InceptionV3 demonstrated a slightly lower performance compared to the top-performing models, with an accuracy of 93%, precision of 93%, recall of 91%, and an F1-score of 92%. ResNet50 showed relatively lower performance with an accuracy of 77%, precision of 78%, recall of 76%, and an F1-score of 76%. These metrics demonstrate the robustness and efficacy of the proposed method for brain tumor detection.
- 25 May 2025
- Published by Springer Nature
- p. 129-150
Lightweight Multi-tier IDS for UAV Networks: Enhancing UAV Zero-Day Attack Detection with Honeypot Threat Intelligence
- 25 May 2025
- Published by Springer Nature
- p. 129-150
Lightweight Multi-tier IDS for UAV Networks: Enhancing UAV Zero-Day Attack Detection with Honeypot Threat Intelligence
- 25 May 2025
- Published by Springer Nature
- p. 153-167
Reducing False Positives in Intrusion Detection System Alerts: A Novel Aggregation and Correlation Model
- 25 May 2025
- Published by Springer Nature
- p. 153-167
Reducing False Positives in Intrusion Detection System Alerts: A Novel Aggregation and Correlation Model
- Open Access
- 31 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 26142-26156
FLSH: A Framework Leveraging Similarity Hashing for Android Malware and Variant Detection
As technology advances, the frequency and complexity of cyberattacks continue to rise, with Android OS, given its significant market share, becoming a prime target for sophisticated malware attacks. While the core categories of malware remain consistent, minor modifications often allow these variants to evade detection, posing significant challenges for security systems. To address this, various techniques and algorithms have been employed to improve malware detection and classification. In this paper, we focus on leveraging fuzzy hashes to calculate the similarity index between files, aiding in the identification of malicious content within seemingly legitimate APK files. Our research enhances the accuracy and reliability of fuzzy hashes, particularly for static features, in detecting Android malware and its variants. Unlike traditional approaches, our method employs a distinctive static feature-based fuzzy hashing technique. We conducted experiments on a dataset of 2000 APK files, including both benign and malicious samples, and classified malware into six categories trojan, adware, spyware, virus, downloader, and hacktool. The results showed a significant improvement in precision, recall, and F-measure, achieving an overall accuracy of 96.67%, without relying on complex machine learning or deep learning methods.
- 31 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 26142-26156
Open AccessFLSH: A Framework Leveraging Similarity Hashing for Android Malware and Variant Detection
As technology advances, the frequency and complexity of cyberattacks continue to rise, with Android OS, given its significant market share, becoming a prime target for sophisticated malware attacks. While the core categories of malware remain consistent, minor modifications often allow these variants to evade detection, posing significant challenges for security systems. To address this, various techniques and algorithms have been employed to improve malware detection and classification. In this paper, we focus on leveraging fuzzy hashes to calculate the similarity index between files, aiding in the identification of malicious content within seemingly legitimate APK files. Our research enhances the accuracy and reliability of fuzzy hashes, particularly for static features, in detecting Android malware and its variants. Unlike traditional approaches, our method employs a distinctive static feature-based fuzzy hashing technique. We conducted experiments on a dataset of 2000 APK files, including both benign and malicious samples, and classified malware into six categories trojan, adware, spyware, virus, downloader, and hacktool. The results showed a significant improvement in precision, recall, and F-measure, achieving an overall accuracy of 96.67%, without relying on complex machine learning or deep learning methods.
- Open Access
- 13 December 2024
- Published by MDPI AG in Journal of Imaging
- Vol. 10 (12) , 322
MediScan: A Framework of U-Health and Prognostic AI Assessment on Medical Imaging
With technological advancements, remarkable progress has been made with the convergence of health sciences and Artificial Intelligence (AI). Modern health systems are proposed to ease patient diagnostics. However, the challenge is to provide AI-based precautions to patients and doctors for more accurate risk assessment. The proposed healthcare system aims to integrate patients, doctors, laboratories, pharmacies, and administrative personnel use cases and their primary functions onto a single platform. The proposed framework can also process microscopic images, CT scans, X-rays, and MRI to classify malignancy and give doctors a set of AI precautions for patient risk assessment. The proposed framework incorporates various DCNN models for identifying different forms of tumors and fractures in the human body i.e., brain, bones, lungs, kidneys, and skin, and generating precautions with the help of the Fined-Tuned Large Language Model (LLM) i.e., Generative Pretrained Transformer 4 (GPT-4). With enough training data, DCNN can learn highly representative, data-driven, hierarchical image features. The GPT-4 model is selected for generating precautions due to its explanation, reasoning, memory, and accuracy on prior medical assessments and research studies. Classification models are evaluated by classification report (i.e., Recall, Precision, F1 Score, Support, Accuracy, and Macro and Weighted Average) and confusion matrix and have shown robust performance compared to the conventional schemes.
- 13 December 2024
- Published by MDPI AG in Journal of Imaging
- Vol. 10 (12) , 322
Open AccessMediScan: A Framework of U-Health and Prognostic AI Assessment on Medical Imaging
With technological advancements, remarkable progress has been made with the convergence of health sciences and Artificial Intelligence (AI). Modern health systems are proposed to ease patient diagnostics. However, the challenge is to provide AI-based precautions to patients and doctors for more accurate risk assessment. The proposed healthcare system aims to integrate patients, doctors, laboratories, pharmacies, and administrative personnel use cases and their primary functions onto a single platform. The proposed framework can also process microscopic images, CT scans, X-rays, and MRI to classify malignancy and give doctors a set of AI precautions for patient risk assessment. The proposed framework incorporates various DCNN models for identifying different forms of tumors and fractures in the human body i.e., brain, bones, lungs, kidneys, and skin, and generating precautions with the help of the Fined-Tuned Large Language Model (LLM) i.e., Generative Pretrained Transformer 4 (GPT-4). With enough training data, DCNN can learn highly representative, data-driven, hierarchical image features. The GPT-4 model is selected for generating precautions due to its explanation, reasoning, memory, and accuracy on prior medical assessments and research studies. Classification models are evaluated by classification report (i.e., Recall, Precision, F1 Score, Support, Accuracy, and Macro and Weighted Average) and confusion matrix and have shown robust performance compared to the conventional schemes.
- Open Access
Seeing Beyond Noise: Improving Cryptocurrency Forecasting with Linear Bias Correction
Cryptocurrency is recognized as a leading digital currency by its peer-to-peer transfer capabilities and secure features. Accurately forecasting cryptocurrency price trends holds substantial significance for investors and traders, as they inform critical decisions regarding the acquisition, divestment, or retention of cryptocurrencies, guided by expectations of value, risk assessment, and potential returns. This study also aims to identify a resourceful technique to efficiently forecast prices of cryptocurrencies such as Bitcoin (BTC), Binance (BNB), Ripple (XRP), and Tether (USDT) using optimal data-driven models (LSTM, GRU, and BiLSTM models) using bias correction. The proposed methodology includes collecting cryptocurrency data and precious metal data from Coindesk and BullionVault, respectively, and then finding the optimal model input combination for each cryptocurrency by lag adjustment and correlating feature selection. Hyperparameter tuning was performed by trial-and-error technique, and an early stopping function was applied to minimize time and space complexity. Bias correction (BC) is applied to model-forecasted price trends to reduce errors in evaluation and to enhance accuracy by adjusting model outputs to reduce prediction bias, providing a refined alternative to traditional unadjusted deep learning methods. GRU-BC outperformed other models in forecasting Bitcoin (with MAE 25.291, RMSE 31.266, MAPE 2.999) and USDT (with MAE 0.0006, RMSE 0.0012, MAPE 0.0622) price trends, while BiLSTM-BC was superior in predicting XRP (with MAE 0.0129, RMSE 0.0171, MAPE 2.9013) and BNB (with MAE 2.2759, RMSE 2.8357, MAPE 1.9785) market price flow.
Open AccessSeeing Beyond Noise: Improving Cryptocurrency Forecasting with Linear Bias Correction
Cryptocurrency is recognized as a leading digital currency by its peer-to-peer transfer capabilities and secure features. Accurately forecasting cryptocurrency price trends holds substantial significance for investors and traders, as they inform critical decisions regarding the acquisition, divestment, or retention of cryptocurrencies, guided by expectations of value, risk assessment, and potential returns. This study also aims to identify a resourceful technique to efficiently forecast prices of cryptocurrencies such as Bitcoin (BTC), Binance (BNB), Ripple (XRP), and Tether (USDT) using optimal data-driven models (LSTM, GRU, and BiLSTM models) using bias correction. The proposed methodology includes collecting cryptocurrency data and precious metal data from Coindesk and BullionVault, respectively, and then finding the optimal model input combination for each cryptocurrency by lag adjustment and correlating feature selection. Hyperparameter tuning was performed by trial-and-error technique, and an early stopping function was applied to minimize time and space complexity. Bias correction (BC) is applied to model-forecasted price trends to reduce errors in evaluation and to enhance accuracy by adjusting model outputs to reduce prediction bias, providing a refined alternative to traditional unadjusted deep learning methods. GRU-BC outperformed other models in forecasting Bitcoin (with MAE 25.291, RMSE 31.266, MAPE 2.999) and USDT (with MAE 0.0006, RMSE 0.0012, MAPE 0.0622) price trends, while BiLSTM-BC was superior in predicting XRP (with MAE 0.0129, RMSE 0.0171, MAPE 2.9013) and BNB (with MAE 2.2759, RMSE 2.8357, MAPE 1.9785) market price flow.