Dimensionality reduction algorithms have been applied widely in computer vision, medical image processing, video processing, face recognition, image retrieval etc. However, some problems need to be fixed. First of all...
详细信息
Dimensionality reduction algorithms have been applied widely in computer vision, medical image processing, video processing, face recognition, image retrieval etc. However, some problems need to be fixed. First of all, how to get the domain size and intrinsic dimension is first primary problem to be fixed, which is seriously restricted the rapid development. Traditonal method used the K-Nearest Neighbor to search the neighborhoods of each image sample. But it costs much time sometimes. In this paper, we aim at the problem of finding intrinsic structure and domain size automatically for high dimensional image data. We present a new technique which can get intrinsic dimension and neighborhood size automatically and adaptively. The algorithm can be used for extracting local features from images. Firstly, we made linear reconstruction based on the nearest neighbor distance for image feature extraction, and optimize the distribution on whole manifold, then we get expression function in which variable is the optimal linear reconstruction for locally low dimensional feature. Lastly, minimiing the variance of the function is to get aotomatic selection strategy. Experiments show that the algorithm is not only simple but also high matching rate and low computational complexity.
Rule extraction is a main goal for rough set theory. This paper mainly constructs a new algorithm (LBRM Algorithm) for rule extraction based on rough membership. The confidence principle is established based on rough ...
详细信息
IoT (Internet of Things) bridges the physical world and information space. IoT services are environment sensitive and event-driven. The new IoT service architecture should adapt to these features. This paper analyses ...
详细信息
IoT (Internet of Things) bridges the physical world and information space. IoT services are environment sensitive and event-driven. The new IoT service architecture should adapt to these features. This paper analyses IoT sensing service characteristics and proposes the future services architecture. It is focused on the middleware architecture and the interface presentation technology. In the middleware layer, the traditional SOA architecture is insufficient in the real-time response and parallel process of services execution, this paper proposes that the new sensing servicesystem based on EDSOA (Event Driven SOA) architecture to support real-time, event-driven, and active service execution. At presentation layer, this paper presents the new IoT browser features including using augmented reality technology to input and output, and realize the superposition presentation of the physical world and abstract information. Through a use case and proof-of-concept implementation-road manhole covers monitoring system - we verify the feasibility of the proposed ideas and frameworks.
Traditional intrusion detection classification based methods could not tackle the abnormal events in a changing network environment, because those methods need lots of labeled data for training prediction. On the othe...
详细信息
Building simulation models play a vital role in optimal building climate control, energy audit, fault detection and diagnosis, continuous commissioning, and planning. Real system parameters are often unknown or partia...
详细信息
This paper derives a more accurate sampled-data model than the Yuz and Goodwin type model for nonlinear systems in the case of the relative degree two, and analyzes the sampling zero dynamics of the sampled-data model...
详细信息
ISBN:
(纸本)9781467355322
This paper derives a more accurate sampled-data model than the Yuz and Goodwin type model for nonlinear systems in the case of the relative degree two, and analyzes the sampling zero dynamics of the sampled-data model to show a condition which assures the stability of the sampling zero dynamics of the proposed model. It is a nature extension of Ishitobi et al.'s result from a single-input single-output (SISO) nonlinear systems to a multi-input multi-output (MIMO) nonlinear systems.
This work intends to schedule a dual-arm cluster tool for the atomic layer deposition (ALD) process with wafer residency time constraints. ALD is a typical wafer revisiting process. Based on the analysis of system pro...
详细信息
A novel alive entropy-based detection approach was proposed, which detects the abnormal network traffic based on the values of alive entropies. The alive entropies calculated based on the NetFlow data coming from the ...
详细信息
A novel alive entropy-based detection approach was proposed, which detects the abnormal network traffic based on the values of alive entropies. The alive entropies calculated based on the NetFlow data coming from the network traffic of input and output of a whole system, which is essentially a monitored network. In order to decrease false positive rate of abnormal network traffic, different scales are selected to compute the values of alive entropies in different sizes of network traffic. With the low false positive rate of abnormal network traffic, the abnormal network traffic can be effectively detected. Experiments carried out on a real campus network were used to evaluate the effectiveness of the proposed approach. A comparative study illustrates that the proposed approach may easily detect the abnormal network traffic with random characteristics in comparison with some "conventional" approaches reported in the literatures.
This paper proposes a novel method to optimizing H.264 decoding on multi-core processors by exploiting color component level parallelism. Compared with sequential decoding and macro-block level parallelism, the propos...
详细信息
In this paper, we propose a deep packet inspection system based on the MapReduce. The MapReduce which is a parallel distributed programming model developed by Google applies the technology of deep packet inspection. N...
详细信息
暂无评论