42 research outputs found
Robust Optimization of MobileNet for Plant Disease Classification with Fine Tuned Parameters
No full text
A Cutting-Edge Hybrid Approach for Precise COVID-19 Detection using Deep Learning
No full textThe early detection of COVID-19 is essential for decision-makers to develop effective containment and treatment plans. Traditionally, researchers interpret computer tomography (CT) scans or X-ray images in order to diagnose this disease. This study aims to demonstrate that deep learning models can be applied to three common medical imaging modes: X-rays, ultrasounds, and CT scans. This study employs and enhances four convolutional neural networks for coronavirus detection, including DenseNet121, ResNet101V2, NASNetMobile, and MobileNetV2. In this study, two main experiments were carried out. In the first experiment, a model was developed by combining imagery data to detect this virus. In order to determine which model performed the best, separate models were trained using different datasets in the second experiment. Because there were only so many photos accessible, data augmentation techniques were used to enhance the amount artificially. The results indicate that the proposed models effectively accomplished the task of classifying COVID-19. The accuracy rates achieved by the combined model, utilizing DenseNet121, ResNet101V2, NASNetMobile, and MobileNetV2, were 88.21%, 93.02%, and 88.89% respectively. When using the combined imaging dataset, the CNN model employing ResNet101v2 exhibited superior accuracy compared to NASNetMobile, DenseNet121, and MobileNetV2 models
A Cutting-Edge Hybrid Approach for Precise COVID-19 Detection using Deep Learning
No full textThe early detection of COVID-19 is essential for decision-makers to develop effective containment and treatment plans. Traditionally, researchers interpret computer tomography (CT) scans or X-ray images in order to diagnose this disease. This study aims to demonstrate that deep learning models can be applied to three common medical imaging modes: X-rays, ultrasounds, and CT scans. This study employs and enhances four convolutional neural networks for coronavirus detection, including DenseNet121, ResNet101V2, NASNetMobile, and MobileNetV2. In this study, two main experiments were carried out. In the first experiment, a model was developed by combining imagery data to detect this virus. In order to determine which model performed the best, separate models were trained using different datasets in the second experiment. Because there were only so many photos accessible, data augmentation techniques were used to enhance the amount artificially. The results indicate that the proposed models effectively accomplished the task of classifying COVID-19. The accuracy rates achieved by the combined model, utilizing DenseNet121, ResNet101V2, NASNetMobile, and MobileNetV2, were 88.21%, 93.02%, and 88.89% respectively. When using the combined imaging dataset, the CNN model employing ResNet101v2 exhibited superior accuracy compared to NASNetMobile, DenseNet121, and MobileNetV2 models.
Manuscript received: 23 Jan 2024 | Revised: 26 Feb 2024 | Accepted: 25 Mar 2024 | Published: : 30 Apr 202
Comparative Analysis of Machine Learning Algorithms for Classification of Environmental Sounds and Fall Detection
Get PDFIn recent years, number of elderly people in population has been increased because of the rapid advancements in the medical field, which make it necessary to take care of old people. Accidental fall incidents are life-threatening and can lead to the death of a person if first aid is not given to the injured person. Immediate response and medical assistance are necessary in case of accidental fall incidents to elderly people. The research community explored various fall detection systems to early detect fall incidents, however, still there exist numerous limitations of the systems such as using expensive sensors, wearable sensors that are hard to wear all the time, camera violates the privacy of person, and computational complexity. In order to address the above-mentioned limitations of the existing systems, we proposed a novel set of integrated features that consist of melcepstral coefficients, gammatone cepstral coefficients, and spectral skewness. We employed a decision tree for the classification performance of both binary problems and multi-class problems. We obtained an accuracy of 91.39%, precision of 96.19%, recall of 91.81%, and F1-score of 93.95%. Moreover, we compared our method with existing state-of-the-art methods and the results of our method are higher than other methods. Experimental results demonstrate that our method is reliable for use in medical centers, nursing houses, old houses, and health care provisions.
Full Tex
Prediction of fatigue crack growth rate in aircraft aluminum alloys using radial basis function neural network
No full textTraining Deep Neural Networks with Novel Metaheuristic Algorithms for Fatigue Crack Growth Prediction in Aluminum Aircraft Alloys
Get PDFFatigue cracks are a major defect in metal alloys, and specifically, their study poses defect evaluation challenges in aluminum aircraft alloys. Existing inline inspection tools exhibit measurement uncertainties. The physical-based methods for crack growth prediction utilize stress analysis models and the crack growth model governed by Paris’ law. These models, when utilized for long-term crack growth prediction, yield sub-optimum solutions and pose several technical limitations to the prediction problems. The metaheuristic optimization algorithms in this study have been conducted in accordance with neural networks to accurately forecast the crack growth rates in aluminum alloys. Through experimental data, the performance of the hybrid metaheuristic optimization–neural networks has been tested. A dynamic Levy flight function has been incorporated with a chimp optimization algorithm to accurately train the deep neural network. The performance of the proposed predictive model has been tested using 7055 T7511 and 6013 T651 alloys against four competing techniques. Results show the proposed predictive model achieves lower correlation error, least relative error, mean absolute error, and root mean square error values while shortening the run time by 11.28%. It is evident through experimental study and statistical analysis that the crack length and growth rates are predicted with high fidelity and very high resolution
FISCAL DECENTRALISATION AND POLITICAL ECONOMY OF POVERTY REDUCTION: THEORY AND EVIDENCE FROM PAKISTAN
Get PDFThis thesis explores the relationship between fiscal decentralisation and poverty. The thesis consists of four parts. First part reviews the related literature addressing different aspects of fiscal decentralisation and poverty and highlighting the research gap that this thesis intends to address. It also explains the possible channels through which fiscal decentralisation potentially affects poverty. Second part describes the political economy, fiscal decentralisation and poverty in Pakistan. It underlines that fiscal policy decisions in Pakistan are made to reflect many vested interest groups and institutions that may be failed to provide basic social services. Additionally, it discusses the development of federalism and fiscal decentralisation in Pakistan and shows that how the vertical and horizontal resource distribution affect the social and economic development of the provinces. This part also discusses various approaches, measurements and trends of poverty in Pakistan. Third part presents a systematic relationship between fiscal decentralisation and poverty both theoretically and empirically. The theoretical framework implies that if the federal transfer rate is larger, then the decentralisation measure will be greater. Since a larger federal transfer rate reduces poverty, poverty and expenditure decentralisation are expected to be negatively related. In addition to the model, there is an extensive empirical study on Pakistan to look at the impact of fiscal decentralisation on poverty besides investigating the potential channels through pro-poor sectoral outcomes. Ordinary Least Squared, Fixed and Radom Effect Models and Generalised Method of Moment Instrumental Variables methodology is used on simple time series as well as panel datasets covering four provinces of Pakistan over the period from 1975 to 2009. The empirical results suggest a strong relationship between expenditure decentralisation and poverty – proxy alternatively by headcount poverty, poverty gap, severity of poverty and the human development index. Both rural and urban poverty reduction have statistically significant relationship with expenditure decentralisation. The results also reveal that decentralisation improves pro-poor sectoral outcomes of education, health and agriculture that consequently affect poverty.
The last part illustrates the effectiveness of the devolution reforms by transferring fiscal, political and administrative authorities to local governments on certain social and economic sectors that are believed to be pro-poor. The evidence shows that the devolution significantly changes the size and magnitude of investment on many social and economic sectors. In all provinces, the investment increases in sectors such as education, healthcare, agriculture, water management, water supply and sanitation, rural development and the civil work. Since these services are strongly associated with local needs, it is reasonable to conclude that the devolution implicitly enhances the living standard of the local communities, especially the poor
