Syed Akbar Ali Shah Akbar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Syed Akbar Ali Shah Akbar has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
Ghulam Ali
Information Technology Centre, Sindh Agriculture University Tando Jam Pakistan
1 shared publicationGulsher Laghari
Institute of Mathematics and Computer Science (IMCS), University of Sindh, Jamshoro
1 shared publication- Gulsher Laghari
Institute of Mathematics and Computer Science, University of Sindh, Jamshoro, Pakistan
1 shared publication Imtiaz Ali Korejo
Nstitute of Mathematics and Computer Science (IMCS), University of Sindh, Jamshoro
1 shared publicationKamran Brohi
Institute of Mathematics and Computer Science (IMCS), University of Sindh, Jamshoro
1 shared publication- Shazia Samoon
Institute of Mathematics and Computer Science, University of Sindh, Jamshoro, Pakistan
1 shared publication
Latest Publications
- Open Access
- 1 August 2025
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 13 (2) , 36-49
HAR-AttenNet: Multi-Head Transformer for Precise Human Activity Recognition Using Wearable Devices
Human Activity Recognition has become inevitable in wearable technology in healthcare and fitness tracking applications. Precise recognition of daily routine human activities using wearable sensors remains a challenging task due to sensor modality, placement, and environmental factors. Indeed, Machine Learning and Deep Learning models have become robust in human activity recognition, yet these face numerous challenges in precisely recognizing the daily human activities. One major challenge is the variability in sensors readings where the same activity may lead to different sensor readings, intra-class variability, as different people perform the same activity differently or use different devices. To address this, we propose a multi-head transformer model with a multi-head attention mechanism that explicitly handles the intra-class variability. We demonstrate that the multi-head Transformer model exhibits enhanced robustness and performs better even in the face of such variability, although variability is a major problem in the popular PAMAP2 dataset, since it directly impacts the performance of deep learning models.
- 1 August 2025
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 13 (2) , 36-49
Open AccessHAR-AttenNet: Multi-Head Transformer for Precise Human Activity Recognition Using Wearable Devices
Human Activity Recognition has become inevitable in wearable technology in healthcare and fitness tracking applications. Precise recognition of daily routine human activities using wearable sensors remains a challenging task due to sensor modality, placement, and environmental factors. Indeed, Machine Learning and Deep Learning models have become robust in human activity recognition, yet these face numerous challenges in precisely recognizing the daily human activities. One major challenge is the variability in sensors readings where the same activity may lead to different sensor readings, intra-class variability, as different people perform the same activity differently or use different devices. To address this, we propose a multi-head transformer model with a multi-head attention mechanism that explicitly handles the intra-class variability. We demonstrate that the multi-head Transformer model exhibits enhanced robustness and performs better even in the face of such variability, although variability is a major problem in the popular PAMAP2 dataset, since it directly impacts the performance of deep learning models.
- Open Access
- 19 March 2025
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 13 (1) , 140-160
A Review on Cardiovascular Diseases Risk Prediction Approches Based on Machine Learning and Evolutionary Algorithms
Across the world, 17 million people die from heart disease each year. Heart-related diseases were the main cause for about 19% percent of deaths in Pakistan in 2016, the same has since now risen to 29%. As per most recent WHO statistics regarding prevalence of heart attacks in Pakistan, approximately more than two hundred thousand (200000) persons died in Pakistan in 2020 due to coronary heart disease, making up 16.49 percent of all fatalities. With a death rate of 193.56 per 100,000 inhabitants, Pakistan is ranked at number 30 in the world. Rising death rate due to heart disease can be minimized through detection at early stage. Different data mining approaches have made early detection of cardiac disease possible. Certain datasets are being used to retrieve useful information. Several machine learning techniques / models have been proved to be the most effective, accurate and profitable to detect cardiovascular disease at an early stage. However, the approach of machine learning and Genetic Algorithm (GA) with feature selection may aid in lowering the computational complexity of GA and increasing the effectiveness of its search for ideal solutions. Hence, there is dire need to apply such hybrid approach to get much more effective and accurate results. The goal of this survey paper is to review different papers related to CVD prediction at early stage by applying hybrid approach of ML Techniques with Genetic Algorithm. Moreover, the results obtained by the authors in reviewed papers are also examined. In the end, this survey will showcase the importance of the hybrid approach in improving accuracy of ML results.
- 19 March 2025
- Published by VFAST Research Platform in VAWKUM Transactions on Computer Sciences
- Vol. 13 (1) , 140-160
Open AccessA Review on Cardiovascular Diseases Risk Prediction Approches Based on Machine Learning and Evolutionary Algorithms
Across the world, 17 million people die from heart disease each year. Heart-related diseases were the main cause for about 19% percent of deaths in Pakistan in 2016, the same has since now risen to 29%. As per most recent WHO statistics regarding prevalence of heart attacks in Pakistan, approximately more than two hundred thousand (200000) persons died in Pakistan in 2020 due to coronary heart disease, making up 16.49 percent of all fatalities. With a death rate of 193.56 per 100,000 inhabitants, Pakistan is ranked at number 30 in the world. Rising death rate due to heart disease can be minimized through detection at early stage. Different data mining approaches have made early detection of cardiac disease possible. Certain datasets are being used to retrieve useful information. Several machine learning techniques / models have been proved to be the most effective, accurate and profitable to detect cardiovascular disease at an early stage. However, the approach of machine learning and Genetic Algorithm (GA) with feature selection may aid in lowering the computational complexity of GA and increasing the effectiveness of its search for ideal solutions. Hence, there is dire need to apply such hybrid approach to get much more effective and accurate results. The goal of this survey paper is to review different papers related to CVD prediction at early stage by applying hybrid approach of ML Techniques with Genetic Algorithm. Moreover, the results obtained by the authors in reviewed papers are also examined. In the end, this survey will showcase the importance of the hybrid approach in improving accuracy of ML results.