Syed Atir Raza Shirazi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Syed Atir Raza Shirazi has a total of 18 co-authors. Here are the top co-authors by number of shared publications:
- Nafeesa Yousaf
Department of Computer Science,School of Systems and Technology, University of Management and Technology, Lahore, Pakistan
2 shared publications - Rabia Khan
Department of Computer Science, Lahore Garrison University, Lahore Pakistan
2 shared publications - Agha Wafa Abbas1 shared publication
Anum Naseem
Department of Computer Science, Lahore Garrison University, Lahore Pakistan
1 shared publication- Ayesha Kamran
Department of Computer Science, School of Systems and Technology, University of Management and Technology, Lahore 54000, Pakistan
1 shared publication - Bilal Hassan
Faculty of Engineering & Environment, Northumbria University, London Campus, London E1 7HT, UK
1 shared publication
Latest Publications
- 23 February 2026
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 14 (1) , 53-65
A Digital Twin-Inspired Hybrid Variational Quantum Reinforcement Learning Framework for Heart Disease Risk Prediction
Despite many advancements in the medicine domain heart disease mortality rates continue climb that is signailing an urgent need for more advanced risk models. In this work we have tackled this challenge by introducing a hybrid framework that combines digital twin modeling with variational quantum reinforcement learning(VQRL). By developing patient-specific digital twins from raw patient clinical data, we are able to simulate unique health states within a specilized reinforcement learning environment. To optimize the diganostic decision-making we leveraged a variational quantum network. This setup utilizes a hybrid quantum-classical optimization loop, making it well suited for the constraints of modern near-term quantum hardware. When put to the test on the "Cleveland Heart Disease dataset" the proposed framework delivered impressive results: achieving a 94.2% accuracy, and an MCC of 0.82. Beyond just high scores in precision 90.4%, and recall 92.2%. Our framework demonstrated remarkable stability and an ability to generalize during the training phase. Ultimately, this fusion of digital twins and quantum computing offers a fresh, scalable path forward for clinical decision making.
- 23 February 2026
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 14 (1) , 53-65
A Digital Twin-Inspired Hybrid Variational Quantum Reinforcement Learning Framework for Heart Disease Risk Prediction
Despite many advancements in the medicine domain heart disease mortality rates continue climb that is signailing an urgent need for more advanced risk models. In this work we have tackled this challenge by introducing a hybrid framework that combines digital twin modeling with variational quantum reinforcement learning(VQRL). By developing patient-specific digital twins from raw patient clinical data, we are able to simulate unique health states within a specilized reinforcement learning environment. To optimize the diganostic decision-making we leveraged a variational quantum network. This setup utilizes a hybrid quantum-classical optimization loop, making it well suited for the constraints of modern near-term quantum hardware. When put to the test on the "Cleveland Heart Disease dataset" the proposed framework delivered impressive results: achieving a 94.2% accuracy, and an MCC of 0.82. Beyond just high scores in precision 90.4%, and recall 92.2%. Our framework demonstrated remarkable stability and an ability to generalize during the training phase. Ultimately, this fusion of digital twins and quantum computing offers a fresh, scalable path forward for clinical decision making.
- Open Access
- 31 July 2025
- Published by MDPI AG in Journal of Imaging
- Vol. 11 (8) , 256
Advancing Early Blight Detection in Potato Leaves Through ZeroShot Learning
Potatoes are one of the world’s most widely cultivated crops, but their yield is coming under mounting pressure from early blight, a fungal disease caused by Alternaria solani. Early detection and accurate identification are key to effective disease management and yield protection. This paper introduces a novel deep learning framework called ZeroShot CNN, which integrates convolutional neural networks (CNNs) and ZeroShot Learning (ZSL) for the efficient classification of seen and unseen disease classes. The model utilizes convolutional layers for feature extraction and employs semantic embedding techniques to identify previously untrained classes. Implemented on the Kaggle potato disease dataset, ZeroShot CNN achieved 98.50% accuracy for seen categories and 99.91% accuracy for unseen categories, outperforming conventional methods. The hybrid approach demonstrated superior generalization, providing a scalable, real-time solution for detecting agricultural diseases. The success of this solution validates the potential in harnessing deep learning and ZeroShot inference to transform plant pathology and crop protection practices.
- 31 July 2025
- Published by MDPI AG in Journal of Imaging
- Vol. 11 (8) , 256
Open AccessAdvancing Early Blight Detection in Potato Leaves Through ZeroShot Learning
Potatoes are one of the world’s most widely cultivated crops, but their yield is coming under mounting pressure from early blight, a fungal disease caused by Alternaria solani. Early detection and accurate identification are key to effective disease management and yield protection. This paper introduces a novel deep learning framework called ZeroShot CNN, which integrates convolutional neural networks (CNNs) and ZeroShot Learning (ZSL) for the efficient classification of seen and unseen disease classes. The model utilizes convolutional layers for feature extraction and employs semantic embedding techniques to identify previously untrained classes. Implemented on the Kaggle potato disease dataset, ZeroShot CNN achieved 98.50% accuracy for seen categories and 99.91% accuracy for unseen categories, outperforming conventional methods. The hybrid approach demonstrated superior generalization, providing a scalable, real-time solution for detecting agricultural diseases. The success of this solution validates the potential in harnessing deep learning and ZeroShot inference to transform plant pathology and crop protection practices.
- Open Access
- 19 March 2025
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 13 (1) , 68-81
RECENT ADVANCES IN MACHINE LEARNING MODELS FOR ANTIVIRAL PEPTIDE PREDICTION
Viral diseases are widespread and their impact is expressed in millions of cases of infection and mortality around the world, chronic viral diseases include COVID-19, HIV, and hepatitis. To prevent and treat these viral infections, novel agents, and antiviral peptides (AVPs) have been developed. Thus, identifying AVPs is crucial because these pieces of information are invaluable throughout the entire pharmaceutical industry and other sciences. This has been done experimentally and computationally but the need for better accurate and efficient predictors persists. This research also reviews current AVP predictors, including the datasets employed, feature representation methods, classification algorithms, and assessment criteria. In our paper, we discuss the weaknesses of the existing techniques, overview the most efficient strategies, and evaluate the benefits and drawbacks of the classifiers. Furthermore, some directions for future work are discussed in detail including enhanced feature representation techniques, selection of features, and classification approaches. The following advancements are targeted at improving the effectiveness of the algorithms used for the correct prediction of AVPs, which makes it possible to develop new antiviral drugs more successfully.
- 19 March 2025
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 13 (1) , 68-81
Open AccessRECENT ADVANCES IN MACHINE LEARNING MODELS FOR ANTIVIRAL PEPTIDE PREDICTION
Viral diseases are widespread and their impact is expressed in millions of cases of infection and mortality around the world, chronic viral diseases include COVID-19, HIV, and hepatitis. To prevent and treat these viral infections, novel agents, and antiviral peptides (AVPs) have been developed. Thus, identifying AVPs is crucial because these pieces of information are invaluable throughout the entire pharmaceutical industry and other sciences. This has been done experimentally and computationally but the need for better accurate and efficient predictors persists. This research also reviews current AVP predictors, including the datasets employed, feature representation methods, classification algorithms, and assessment criteria. In our paper, we discuss the weaknesses of the existing techniques, overview the most efficient strategies, and evaluate the benefits and drawbacks of the classifiers. Furthermore, some directions for future work are discussed in detail including enhanced feature representation techniques, selection of features, and classification approaches. The following advancements are targeted at improving the effectiveness of the algorithms used for the correct prediction of AVPs, which makes it possible to develop new antiviral drugs more successfully.
- Open Access
- 27 December 2024
- Published by Sir Syed University of Engineering and Technology in Sir Syed University Research Journal of Engineering & Technology
- Vol. 14 (2) , 19-28
A Novel Approach for BMI and Nutrition Monitoring of Dairy Farm Using IoT
In this research article, an Internet of Things (IoT)-based sensor system is used to measure the weight of cattle. This system is composed of a load cell sensor, a microcontroller, and a wireless communication module. So the load cell sensor is placed under cow feet and a load cell sensor that will be measuring weight, will send this data to the microcontroller side where all unprocessed (just measured) cumulative loads of cattle become processed, filter out unnecessary results later on the wireless medium, transmission can be seen at server storage which implements cloud base. Specifically, we are designing the system to be cheap and low power (simple), but also simple to deploy over a variety of agricultural systems. Using the data to determine cattle growth, health, and feeding patterns can then allow the farmers to make a highly informed decision on how best to maximize their operations. This device allows farmers to track the health, and behavior of their cattle remotely alerting any unusual activity or potential health issues early on. Farmers can analyze the data generated by the system to provide better feeding and management of their livestock, ensuring that adequate nutrition and care are received. The other developed IoT technologies could also be combined with the system to create a comprehensive solution for precision livestock production. In conclusion, this proposed system could change the cattle industry by giving farmers crucial health and well-being information about their livestock. It can also help to make livestock farming more efficient and sustainable by enabling early detection of health problems as well as optimizing feeding, and management practices. Overall, there will be an advantage of using this recommended approach because it can improve the condition of animals, increase farm productivity, and promote environmentally friendly livestock production practices.
- 27 December 2024
- Published by Sir Syed University of Engineering and Technology in Sir Syed University Research Journal of Engineering & Technology
- Vol. 14 (2) , 19-28
Open AccessA Novel Approach for BMI and Nutrition Monitoring of Dairy Farm Using IoT
In this research article, an Internet of Things (IoT)-based sensor system is used to measure the weight of cattle. This system is composed of a load cell sensor, a microcontroller, and a wireless communication module. So the load cell sensor is placed under cow feet and a load cell sensor that will be measuring weight, will send this data to the microcontroller side where all unprocessed (just measured) cumulative loads of cattle become processed, filter out unnecessary results later on the wireless medium, transmission can be seen at server storage which implements cloud base. Specifically, we are designing the system to be cheap and low power (simple), but also simple to deploy over a variety of agricultural systems. Using the data to determine cattle growth, health, and feeding patterns can then allow the farmers to make a highly informed decision on how best to maximize their operations. This device allows farmers to track the health, and behavior of their cattle remotely alerting any unusual activity or potential health issues early on. Farmers can analyze the data generated by the system to provide better feeding and management of their livestock, ensuring that adequate nutrition and care are received. The other developed IoT technologies could also be combined with the system to create a comprehensive solution for precision livestock production. In conclusion, this proposed system could change the cattle industry by giving farmers crucial health and well-being information about their livestock. It can also help to make livestock farming more efficient and sustainable by enabling early detection of health problems as well as optimizing feeding, and management practices. Overall, there will be an advantage of using this recommended approach because it can improve the condition of animals, increase farm productivity, and promote environmentally friendly livestock production practices.
- Open Access
- 16 May 2024
- Published by VFAST Research Platform in VFAST Transactions on Software Engineering
- Vol. 12 (2) , 213-230
Issues, Challenges, and Solutions in Data Acquisition in Virtual and Augmented Reality Environments
This paper looks at some of the challenges associated with data acquisition in VR and AR environments, principally by incorporating the privacy of digital forensic and sensor technology. While VR and AR technologies are mainly seen as providing an immersive experience, they also pose significant challenges in collecting data and protecting data collected in environments for privacy. It will look into advanced sensor technologies of high-resolution cameras, inertial measurement units, and biosensors for data accuracy and efficiency. This further researches the methods of data fusion, in particular, Kalman filtering and machine learning-based fusion. Lastly, the role of edge computing in local data processing to reduce the demands for latency and bandwidth is analyzed to allow for real-time processing. It also discusses privacy-enhancing technologies, such as differential privacy and homomorphic encryption, to ensure the protection of user data while maintaining ethical standards. The present article is aimed at implementing a comprehensive framework integrating these technologies to address both technical and moral problems associated with data acquisition through VR and AR for secure and efficient application in these fields.
- 16 May 2024
- Published by VFAST Research Platform in VFAST Transactions on Software Engineering
- Vol. 12 (2) , 213-230
Open AccessIssues, Challenges, and Solutions in Data Acquisition in Virtual and Augmented Reality Environments
This paper looks at some of the challenges associated with data acquisition in VR and AR environments, principally by incorporating the privacy of digital forensic and sensor technology. While VR and AR technologies are mainly seen as providing an immersive experience, they also pose significant challenges in collecting data and protecting data collected in environments for privacy. It will look into advanced sensor technologies of high-resolution cameras, inertial measurement units, and biosensors for data accuracy and efficiency. This further researches the methods of data fusion, in particular, Kalman filtering and machine learning-based fusion. Lastly, the role of edge computing in local data processing to reduce the demands for latency and bandwidth is analyzed to allow for real-time processing. It also discusses privacy-enhancing technologies, such as differential privacy and homomorphic encryption, to ensure the protection of user data while maintaining ethical standards. The present article is aimed at implementing a comprehensive framework integrating these technologies to address both technical and moral problems associated with data acquisition through VR and AR for secure and efficient application in these fields.