Vishnu Baligodugula

and 3 more

Niveshitha Niveshitha

and 3 more

Smart cities have emerged to tackle life critical challenges that can thwart the overwhelming urbanization process, such as expensive health care, increasing energy demand, traffic jams, and environmental pollution. This paper proposes efficient and high-quality cloud-based machine-learning solutions for safe urban smart city environment. For that, supervised MLbased models, i.e., regression and classification, are developed utilizing cloud-based solutions to forecast high performance in execution time and enhanced quality of the solution in terms of the accuracy of the implemented cloud-based ML solution. To predict AQI, i.e. air quality index, ML models utilize pollutants in the air data sets. The mean absolute error, mean squared error, root means the squared error, R2 score are used to validate and test the designed models. As classification models, we perform the support vector machine and random forest algorithms, which are measured using the accuracy score and confusion matrix. Execution times and accuracy of the developed models are computed and contrasted with the times for the cloud-based versions of these models. The results show that among the regression algorithms, lasso regression has an r2 score of 80 percent, while linear regression has an r2 score of 75 percent. Furthermore, among the classification models, the random forest algorithm performs better with an accuracy of 99 percent than the support vector machine approach with 95 percent accuracy. In conclusion, our findings demonstrate that run-time is minimized when models are executed on a cloud platform compared to a desktop machine. Moreover, the accuracy of our models is maintained with reduced execution time.

Ashutosh Ghimire

and 4 more

Context-aware computing applications depend on embedded hardware systems, utilizing sensors embedded in the hardware to gather real-time data and interact with specialized OS software (firmware) for autonomous processing and analysis of intelligent data. Securing embedded IC hardware systems and ensuring their trustworthiness requires an intelligent approach to effectively detect spontaneous hardware Trojans (HTs) insertions and modification attacks aiming to compromise the system's integrity, potentially leaking sensitive information or causing destruction. Implementing robust and advanced intrusion detection systems against supply chain hardware Trojan to countermeasure and continuous monitoring the behavior of these malicious hardware is essential to enable trust in context-aware computing applications. AI-enabled hardware side-channel analysis, involving power and timing assessments, assists in detection of anomalies that may signify potential Trojans. This paper propose intelligent AI approach utilizing unsupervised machine learning in conjunction with hardware side-channel analysis to eliminate the need for golden data samples and efficiently detect hardware Trojan detection. Employing unsupervised clustering, the methodology not only showcased a superior false positive rate but also demonstrated a comparable accuracy level when compared to supervised counterparts such as the K-Nearest Neighbors (KNN) classifier, Support Vector Machine (SVM), and Gaussian classifier-methods reliant on the availability of golden data for training. Notably, the proposed model exhibited an impressive accuracy rate of 93%, particularly excelling in pinpointing diminutive Trojans triggered by concise events, surpassing the capabilities of preceding techniques. In conclusion, this research advances a groundbreaking paradigm in hardware Trojan detection, accentuating its potential in bolstering the integrity of semiconductor IC supply chains.
Dear Editor: This manuscript is already online in techrxiv (https://doi.org/10.36227/techrxiv.12376427.v1), and the co-authors are not correctly associated with the published preprint so we are submitting this again by associating the co-authors. We have also improved the similarity report, the similarity index is 11% of this submission. Thank you. Abstract: Air pollution is one of the significant causes of mortality and morbidity every year. In recent years, many researchers have focused their attention on the associations of air pollution and health. These studies used two types of data in their studies, i.e., air pollution data and health data. Feature engineering is used to create and optimize air quality and health features. In order to merge these datasets residential address, community/county/block/city and hospital/school address are used. Using residence address or any location becomes a spatial problem when the Air Quality Monitoring (AQM) stations are concentrated in urban areas within the regions and an overlap in the AQM stations in urban areas coverage area, which raises the question that how to associate the patients with the relevant AQM station. Also, in most of the studies the distance of patients to the AQM stations is also not taken into account. In this study, we propose a four-part spatial feature engineering algorithm to find the coordinates for health data, calculate distances with AQM stations and associate health records to the nearest AQM station. Hence, removing the limitations of current air pollution health datasets. The proposed algorithm is applied as a case study in Klang Valley, Malaysia. The results show that the proposed algorithm can generate air pollution health dataset efficiently and the algorithm also provides the radius facility to exclude the patients who are situated far away from the stations.
COVID-19 has stunned the global economy and threatened human life. Due to rapidly emerging fatalities and enormous cases appearing every day, researchers across the globe are producing significant contributions to mitigate this pandemic. Besides the race for discovering a vaccine and treatment for COVID-19, there is utmost focus on flattening the curve by undertaking appropriate measures. The remarkable role of frontline medical practitioners, who are eagerly treating the affected people will be penned in the history books. The efforts of scientists and technologists will be remembered for their extraordinary contributions to assist healthcare professionals and governments in mitigating the threats of COVID-19. Leading technology firms have formed consortiums and research groups, which provide funding and free access to supercomputers for solving complex computational problems to eliminate COVID-19. In this research, we have unveiled five state-of-the-art technologies and their remarkable applications that can be used to mitigate and eliminate the problems of COVID-19. These technologies include Artificial Intelligence (AI), 3D Printing Technology (3DPT), Big Data Analytics (BDA), High Performance Computing (HPC) and Telecommunication Technology (TT). This research investigates the use of technology to encounter COVID-19 and aims to serve as the primary reference for promoting future research as well as developments to produce solutions for COVID-19 using AI, 3DPT, BDA, HPC and TT.