Information explosion is a typical feature of the Big data. Learners can easily find a wide variety of knowledge information online. However, the expansion of information also makes it difficult for learners to retrie...
详细信息
As the emerging technology of Unmanned Aerial Vehicles (UAVs) becomes mature, UAVs are widely used in environmental monitoring, communication and other fields. In view of this, this paper analyzed the task of synergis...
详细信息
Aiming at mode selection in D2D communication, a mode selection strategy based on social-aware was proposed. In addition to physical connection status, users’ social relationships also served as an observation elemen...
详细信息
In real-world scenarios, laser-scanned point clouds often suffer from missing data points due to occlusion and other factors, significantly impacting subsequent tasks. We propose a novel point cloud completion method ...
详细信息
This paper proposes a scene segmentation method based on optimized line detection to address the difficulty of segmenting spatially independent equipment in complex substation scenes. The method first employs the RANS...
详细信息
As an important stage in machine learning pipeline, feature selection techniques are mainly used to improve the generalization performance and training efficiency of machine learning model, but few works have focused ...
详细信息
As an important stage in machine learning pipeline, feature selection techniques are mainly used to improve the generalization performance and training efficiency of machine learning model, but few works have focused on the robustness of machine learning models from the perspective of feature selection when dealing with adversarial attacks. In this paper, we propose the adversarial training (AT) based feature selection framework, i.e. AT based feature selection, to improve the robustness of machine learning model built on the feature selection result, which is inspired by using adversarial training to improve the robustness of deep learning model. AT based feature selection framework is the combination of adversarial training with some traditional feature selection algorithm, which can be divided into AT in-processing and AT post-processing feature selection. On the other hand, stability is also a very important property for feature selection. Then we experimentally analyze the relationship between robustness and stability of AT based feature selection, especially theoretically analyze the stability of $\ell_processing$ regularized AT in-processing feature selection algorithm in two different adversarial training forms. Our experimental results on benchmark data sets show that AT based feature selection algorithm is effective to improve the robustness of machine learning model, however, obtain lower stability than corresponding feature selection model without AT.
With the continuous development of artificial intelligence technology, machine learning in distributed network systems, such as IoVflntemet of Vehicles), will inevitably lead to privacy leakage. At present, there are ...
With the continuous development of artificial intelligence technology, machine learning in distributed network systems, such as IoVflntemet of Vehicles), will inevitably lead to privacy leakage. At present, there are lots of problems in federated learning, differential privacy and other machine learning privacy protection schemes, such as high loss of availability and the need of a fully trusted third-party server. In order to solve these problems, we propose MDPFL(Multiple Differential Privacy based on Federated Learning) algorithm. The algorithm combines with the differential privacy model in each stage of federated learning to solve the problem of curious third-party data collectors obtaining users’ original data in the process of machine learning. Meanwhile, the algorithm does not adopt a decentralized machine learning scheme directly, but uses a double noise adding mode with the existence of the third-party data collectors. We use Laplacian mechanism in the central server and GRR mechanism in local clients to ensure the goal of stability of the effect of the machine ¡earning model. The algorithm is compared with FedAvg, CDP, LDP algorithm in accuracy and loss rate which based on EMNIST data sets. In different global sensitivities, the model training effect is consistent with FedAvg algorithm while comparing with LDP algorithm, the speed of model convergence is improved.
For large-scale multitask wireless sensor networks (LSM-WSNs), the traditional data collection mode could suffer low energy-efficiency on data transmission, since the large-scale multitask scenarios could result in mu...
For large-scale multitask wireless sensor networks (LSM-WSNs), the traditional data collection mode could suffer low energy-efficiency on data transmission, since the large-scale multitask scenarios could result in much higher packet collision probability, especially for harsh environments. Mobile data collection is an efficient data acquisition way to prolong network lifetime for LSM-WSNs. However, the mobile collectors could suffer electricity shortage problem, since the limited battery capacity of any mobile collector could not afford the energy consumption of its long-distance movement and massive data collection in large-scale multitask scenarios. Deploying wireless chargers to supplement the energy of mobile collectors is a feasible solution to electricity shortage problem, but will incur extra charger deployment cost. In this paper, we focus on the problem that how to optimize such charger deployment cost, which is NP-hard. By transforming it into minimum-cost submodular cover problem, we devise an efficient approximation algorithm with a provable approximation ratio. The extensive simulation results reveal that our solution always outperforms the other solutions under whatever configurations.
The increasing numbers of the applications and requirement of cloud computing have made huge power consumption in data centers, which brings the problems of the high cost and resource waste. This problem attracts sign...
详细信息
ISBN:
(数字)9781728143286
ISBN:
(纸本)9781728143293
The increasing numbers of the applications and requirement of cloud computing have made huge power consumption in data centers, which brings the problems of the high cost and resource waste. This problem attracts significant attention from academia and industry. A critical approach to solve this problem is constructing an intelligent energy management system for data centers. Furthermore, an efficient assessment and prediction module of power consumption in data centers is an essential part of the management system. It facilitates cloud service providers to perform workflow scheduling at the minimal cost and energy efficiency management with the requirement of QoS. Since the assessment and prediction of power consumption correlate, this paper presents a multi-granularity approach for power consumption prediction in data centers, which combines multi-task learning with the LSTM network. We first transfer a multi-granularity power prediction problem into a multi-task regression problem to assess and predict the power consumption of data center system maintenance and scheduling operations. Due to the time requirement for workflow and container scheduling, the prediction interval is 30 seconds. Then we propose an efficient long short-term memory network for the multigranularity prediction. The experimental results show our model outperforms other prediction models on the real datasets.
Alzheimer's disease (AD) is one of the most common neurodegenerative diseases. Understanding AD progression can empower the patients in taking proactive care. Mini Mental State Examination (MMSE) and AD Assessment...
详细信息
ISBN:
(数字)9781728185262
ISBN:
(纸本)9781728185279
Alzheimer's disease (AD) is one of the most common neurodegenerative diseases. Understanding AD progression can empower the patients in taking proactive care. Mini Mental State Examination (MMSE) and AD Assessment Scale Cognitive subscale (ADAS-Cog) are two prevailing clinical measures designed to evaluate the AD progression. In this paper, we propose a weakly supervised Temporal Multitask Matrix Completion (TMMC) framework, which combines a novel transductive multitask feature selection scheme, to simultaneously predict AD progression measured by MMSE and ADAS-Cog, and identify related biomarkers trackable of AD progression. Specifically, by treating the prediction of cognitive scores at each time point as a regression task, we first formulate AD progression problem as a standard Multitask Matrix Completion (MMC) model. Secondly, considering the limited number of samples available in this study, we introduce a transductive feature selection scheme to jointly select the task-shared features for multiple time points and the task-specific features for different time points, and thus alleviate the over-fitting defect caused by Small-Sample-Size issue. Thirdly, aiming at the small change of cognitive scores between successive time points for a patient, we employ a temporal regularization scheme to capture the temporal smoothness of cognitive scores. Furthermore, we design an efficient optimization algorithm based on Alternative Minimization and Difference of Convex Programming techniques to solve the proposed TMMC framework. Finally, the extensive experiments performed on real-world Alzheimer's disease dataset demonstrate the effectiveness of our TMMC framework.
暂无评论