Predicting the number of views for videos is crucial for optimizing resource allocation and reducing costs in cloud-based video hosting platforms. In this manuscript, we propose a novel method for view count predictio...
详细信息
Predicting the number of views for videos is crucial for optimizing resource allocation and reducing costs in cloud-based video hosting platforms. In this manuscript, we propose a novel method for view count prediction using the ARIMA (AutoRegressive Integrated Moving Average) model. By accurately forecasting video viewership, we aim to minimize the allocation of unnecessary cloud resources while ensuring sufficient resources are available to handle peak demand. Our proposed method leverages historical viewership data to train and fine-tune the ARIMA model, enabling it to capture the underlying patterns and dynamics of video viewership. Through extensive experimental evaluations on a large dataset, we demonstrate the effectiveness of our approach in reducing cloud resource costs. Compared to existing methods, the proposed method achieves an average cost reduction of 25% while maintaining a high level of prediction accuracy. Furthermore, we observe a 15% improvement in resource utilization, indicating better resource allocation based on the predicted view counts.
A dynamic average consensus (DAC) control scheme is proposed in the presence of con-strained communication, where event-triggered cloud access is presented with a cloud repository for agents. With this framework, agen...
详细信息
A dynamic average consensus (DAC) control scheme is proposed in the presence of con-strained communication, where event-triggered cloud access is presented with a cloud repository for agents. With this framework, agents upload and download information from the cloud repository, in which they exchange information with neighbors indirectly. We first propose a triggered condition with a deviation of control signals for reduction of the frequency of communication with the cloud repository. Due to the fact that the neigh-bors' information in the condition can't be obtained in the cloud framework, cloud access scheduling rules, with an adaptive threshold for control signal updating, are then intro-duced via available information at current instants, and agents access the cloud repository in an asynchronous manner to avoid congestion. Available information from the cloud is employed, as well as their own, into the triggered condition for prediction of next access instants, and the triggered consequence is generated. Stability analysis presents that the control scheme in this paper steers the MAS to achieve DAC with an ultimately bounded tracking error and rules out Zeno behaviors. The effectiveness of the control scheme is verified via an example.(c) 2023 Elsevier Inc. All rights reserved.
In the context of the Internet of Things (IoT) era, securing and privatizing IoT-enabled healthcare systems present significant challenges, particularly in ensuring the confidentiality, integrity, and availability of ...
详细信息
ISBN:
(纸本)9783031764585;9783031764592
In the context of the Internet of Things (IoT) era, securing and privatizing IoT-enabled healthcare systems present significant challenges, particularly in ensuring the confidentiality, integrity, and availability of health data exchange. This study explores data fragmentation and employs polynomial and Newton-Gregory's divided difference interpolation techniques for encrypting sensitive health information, such as patient IDs, to enhance data security and utility. The research aims to improve data integrity and ensure end-user availability by fragmenting data. The performance of this methodology is thoroughly evaluated against modern techniques, showing notable superiority in precision, recall, and F-1-score across different correlation index values. Moreover, the study's analysis of time complexity for overhead tasks highlights its efficiency compared to existing technologies. By emphasizing the need for collective efforts in addressing security and privacy concerns, this research contributes to building trust and encouraging the adoption of sophisticated healthcare technologies, paving the way for a secure, data-driven healthcare future.
The three pillars of health data exchange, confidentiality, integrity and availability, pose a significant challenge to the efficiency and robustness of the healthcare ecosystem. By utilising fragmentation, sensitive ...
详细信息
ISBN:
(纸本)9798350310900
The three pillars of health data exchange, confidentiality, integrity and availability, pose a significant challenge to the efficiency and robustness of the healthcare ecosystem. By utilising fragmentation, sensitive attributes dissociate, and thus, data security can be enhanced, and data utility can be improved. Throughout this research, confidentiality was performed by deploying polynomials and Newton-Gregory's divided difference interpolation to enable encryption of confidential data values such as patients' IDs. The fragmentation technique was utilised to achieve integrity, and the utility method enabled end-user availability. Extensive evaluations show that the precision, recall, and F1-score under different values of correlation index. of the proposed methodology outperform state-of-the-art approaches. Also, a time complexity comparison for overhead tasks was implemented between these approaches.
Internet of Vehicles(IoV) is a leading technology of the present era. It has gained huge attention with respect to its implementation in wide variety of domains ranging from traffic safety to infotainment applications...
详细信息
Internet of Vehicles(IoV) is a leading technology of the present era. It has gained huge attention with respect to its implementation in wide variety of domains ranging from traffic safety to infotainment applications. However, Io V can also be extended to healthcare domain, where the patients can be provided healthcare services on-the-fly. We extend this novel concept in this paper and refer it as“Healthcare services on-the-fly”. The concept of game theory has been used among the vehicles to access the healthcare services while traveling. The vehicles act as players in the game and tend to form and split coalitions to access these services. Learning automata(LA) act as the players for interaction with the environment and take appropriate actions based on reward and penalty. Apart from this, Virtual Machine(VM) scheduling algorithm for efficient utilization of resources at cloud level has also been formulated. A stochastic reward net(SRN)-based model is used to represent the coalition formation and splitting with respect to availability of resources at cloud level. The performance of the proposed scheme is evaluated using various performance evaluation metrics. The results obtained prove the effectiveness of the proposed scheme in comparison to the best, first, and random fit schemes.
This paper investigates a multi-agent formation control problem with event-triggered control updates and additive disturbances. The agents communicate only by exchanging information in a cloud repository. The communic...
详细信息
ISBN:
(纸本)9783952426937
This paper investigates a multi-agent formation control problem with event-triggered control updates and additive disturbances. The agents communicate only by exchanging information in a cloud repository. The communication with the cloud is considered a shared and limited resource, and therefore it is used intermittently and asynchronously by the agents. The proposed approach takes advantage of having a shared asynchronous cloud support while guaranteeing a reduced number of communication. More in detail, each agent schedules its own sequence of cloud accesses in order to achieve a coordinated network goal. A control law is given with a criterion for scheduling the control updates recursively. The closed loop scheme is proven to be effective in achieving the control objective and a numerical simulation corroborates the theoretical results.
'TaggerVR' is a work-in-progress immersive VR implementation of the 'Tagger' interactive software tool designed to visualize, characterize, sample and tag geoscientific datasets hosted in local and clo...
详细信息
ISBN:
(纸本)9781467373432
'TaggerVR' is a work-in-progress immersive VR implementation of the 'Tagger' interactive software tool designed to visualize, characterize, sample and tag geoscientific datasets hosted in local and cloud-based repositories via a THREDDS Data Server and OPeNDAP. TaggerVR implements a VR GUI using Human Interface Devices (HID), providing data to the user via an animated interface. This enables interesting features that would otherwise be lost in large datasets to be 'Tagged' for subsequent analysis. A key novel enabler is the link between scientific data formats and the high-performance interactive graphics.
暂无评论