Most of the research on clustering ensemble focuses on designing practical consistency learning algorithms. To solve the problems that the quality of base clusters varies and the low-quality base clusters have an impact on the performance of the clustering ensemble， from the perspective of data mining， the intrinsic connections of data were mined based on the base clusters， and a high-order information fusion algorithm was proposed to represent the connections between data from different dimensions， namely Clustering Ensemble with High-order Consensus learning （HCLCE）. Firstly， each high-order information was fused into a new structured consistency matrix. Then， the obtained multiple consistency matrices were fused together. Finally， multiple information was fused into a consistent result. Experimental results show that LCLCE algorithm has the clustering accuracy improved by an average of 7.22%， and the Normalized Mutual Information （NMI） improved by an average of 9.19% compared with the suboptimal Locally Weighted Evidence Accumulation （LWEA） algorithm. It can be seen that the proposed algorithm can obtain better clustering results compared with clustering ensemble algorithms and using one information alone.

For the cutting stock problem of circular parts which is widely existed in many manufacturing industries， a new parallel genetic algorithm for cutting stock was proposed to maximize the material utilization within a reasonable computing time， namely Parallel Genetic Blanking Algorithm （PGBA）. In PGBA， the material utilization rate of cutting plan was used as the optimization objective function， and the multithread was used to perform the genetic manipulation on multiple subpopulations in parallel. Firstly， a specific individual coding method was designed based on the parallel genetic algorithm， and a heuristic method was used to generate the individuals of population to improve the search ability and efficiency of the algorithm and avoid the premature phenomena. Then， an approximate optimal cutting plan was searched out by adaptive genetic operations with better performance. Finally， the effectiveness of the algorithm was verified by various experiments. The results show that compared with the heuristic algorithm proposed in literature， PGBA takes longer computing time， but has the material utilization rate greatly improved， which can effectively improve the economic benefits of enterprises.

Scientists identify the species of whales based on the shape and the distinctive marks of the whale tails， but the process of recognition by human eyes and manual labeling is very cumbersome. The dataset of whale tail photo has the unbalanced data distribution， and some specific categories in the dataset have very few samples or even one sample. Besides， the samples have small individual differences and contain unknown categories， which leads to the difficulty in automatic labeling of whale identification by image classification. To solve the problem that metric learning is difficult to realize classification under this task， on the basis of Siamese Neural Network （SNN）， the training batches were constructed dynamically by using Linear Assignment Problem （LAP） algorithm in the training process of hard-negative sample mining. Firstly， image feature vectors were extracted from the training samples， and the similarity metric of feature vector was calculated. Then， LAP was used to assign sample pairs to the model， training sample batches were constructed dynamically according to the metric score matrix， and the difficult sample pairs were targeted by trained. Experimental results on a whale tail image dataset with unbalanced data distribution and CUB 200-2001 dataset show that， the proposed algorithm can achieve good results in learning minority classes and classifying fine-grained images.

In most existing multi-manifold face recognition algorithms, the original data with noise are directly processed, but the noisy data often have a negative impact on the accuracy of the algorithm. In order to solve the problem, a Robust Multi-Manifold Discriminant Local Graph Embedding algorithm based on the Maximum Margin Criterion (RMMDLGE/MMC) was proposed. Firstly, a denoising projection was introduced to process the original data for iterative noise reduction, and the purer data were extracted. Secondly, the data image was divided into blocks and a multi-manifold model was established. Thirdly, combined with the idea of maximum margin criterion, an optimal projection matrix was sought to maximize the sample distances on different manifolds while to minimize the sample distances on the same manifold. Finally, the distance from the test sample manifold to the training sample manifold was calculated for classification and identification. The experimental results show that, compared with Multi-Manifold Local Graph Embedding algorithm based on the Maximum Margin Criterion (MLGE/MMC) which performs well, the classification recognition rate of the proposed algorithm is improved by 1.04, 1.28 and 2.13 percentage points respectively on ORL, Yale and FERET database with noise and the classification effect is obviously improved.

The uncertainty of workload in physical hosts easily leads to high overloaded risk and low resource utilization in physical hosts, which will further affect the energy consumption and service quality of data center. In order to solve this problem, a Workload Uncertainty-based Virtual Machine Consolidation (WU-VMC) method was proposed by analyzing the workload records of physical hosts and the historical data of virtual machine resource request. In order to stabilize the workload of each host in the cloud data center, firstly, the workloads of physical hosts were fitted according to resource requests of virtual machines, and the virtual machine matching degree between virtual machines and physical hosts was computed by using gradient descent method. Then, the virtual machines were integrated by using the matching degree to solve the problems such as increased energy consumption and decreased service quality which were caused by uncertain load. The simulation experimental results show that the proposed WU-VMC method can decrease energy consumption and virtual machine migration times of data center, improving the resource utilization and service quality of data center.

Concerning the problem that the workload of hosts in data center cannot maintain long-term stability by executing traditional Virtual Machine Consolidation (VMC), a high efficient Gaussian Mixture Model-based VMC (GMM-VMC) method was proposed. Firstly, to accurately predict the variation trend of workload in hosts, Gaussian Mixture Model (GMM) was used to fit the workload history of hosts. Then, the overload probability of a host was calculated according to the GMM of its workload and resource capacity. Next, the aforementioned overload probability was taken as the criteria to determine whether the host is overloaded or not. Besides, some virtual machines hosted by overloaded hosts which can significantly degrade overload risk and demand less migration time were selected to migrate. At last, these migrated virtual machines were placed in new hosts which have less effect on workload variation after placement estimated by GMM. Using CloudSim toolkit, GMM-VMC method was validated and compared with other methods on energy consumption, Quality of Service (QoS) and efficiency of consolidation. The experimental results show that the GMM-VMC method can degrade energy consumption in data center and improve QoS.

Considering the uncertainty of dynamic workloads in cloud computing, an Virtual Machine (VM) dynamic consolidation method based on adaptive overloaded threshold selection was proposed. In order to make a trade-off between energy efficiency and Quantity of Services (QoS) of data centers, an adaptive overloaded threshold selection problem model based on Markov decision processes was designed. The optimal decision was calculated by solving this problem model, and the overloaded threshold was dynamically adjusted by using the optimal decision according to energy efficiency and QoS of data center. Overloaded threshold was used to predict overloaded hosts and trigger VM migrations. According to the principle of minimum migration time and minimum energy consumption growth, the VM migration strategy under overloaded threshold constraint was given, and the underloaded hosts were switched to sleep mode. Simulation results show that this method can significantly avoid excessive virtual machine migrations and decrease the energy consumption while improving QoS effectively; in addition, it can achieve an ideal balance between QoS and energy consumption of data center.

Protocol state machine can describe the behavior of a protocol, which can help to understand the behavior logic of protocol. Oriented towards text protocols, a statistical method was firstly used to extract the semantic keyword of representative message type, and an adjacency matrix was used to describe the sequential relationship between the message types, based on which the protocol states were labeled and a state transition diagram was built. The experimental results show that the method can accurately describe the sequential relationship between the message types and abstract state machine model accurately.

To achieve automatic analysis for different types of ElectroCardioGraph (ECG), a sequential screening method for maximum value was brought to detect R wave, while Support Vector Machine (SVM) was used to identify arrhythmia heart beats finally. The localization algorithm based on mathematical morphology combined with characteristics of ECG defined R-wave screening interval to avoid threshold selection in traditional algorithm. After R-peaks being positioned, various types of arrhythmia heart beats were extracted with R wave crest as its center and classified by selecting Radial Basis Function (RBF) or SVM. The results of the simulation experiment on the MIT-BIH database files indicate that this algorithm acquired high relevance ratio at 99.36% for ECG with different types of heart beats. After learning, the SVM can effectively identify as many as 4 types, such as atrial premature beat, premature ventricular beat, bundle branch block and normal heart beat, the overall recognition rate is 99.75%.

Local features measuring image clarity are studied. Energy of Laplacian (EOL) is considered as the optimal feature. An novel intelligent image fusion algorithm based on EOL is proposed. A set of registered images is firstly segmented, then local EOLs of segmented image blocks are computed. EOLs are input into neural network and the target vectors are automatically obtained by comparing values of EOLs. Test images are segmented and their EOLs are put into trained network and the rough fusion images are generated. Final fusion results will be obtained by consistence verification. Experimental results demonstrated the good fusion performance on different source images.

Present User Generated Content (UGC) video system mainly adopts the client/server architecture, which can result in huge bandwidth pressure on streaming server. This paper proposed a Peer-to-Peer (P2P) based online short video search policy—FastSearch. It aims to make use of the relevancy relationship between video resources to locate resource, which can improve the sharing efficiency between peers and decrease the consumption of streaming server. The simulation results show that FastSearch has an efficient streaming source peers searching ability and can greatly reduce the bandwidth consumption of streaming server.

A new method of filtering mixed noise based on limited grayscale and Pulse Coupled Neural Network (PCNN) was proposed for an image contaminated by salt and pepper noise and Gaussian noise. First, salt and pepper noise was identified according to the limited grayscale in a detecting window. Then the noise was filtered via mean filter in a filtering window. Subsequently, Gaussian noise was identified by using the time matrix of PCNN. Finally the Gaussian noise was filtered by some different filters based on variable step. The experimental results show that the proposed method has more advantages not only in filtering effects but also in objective evaluation indexes of Peak Signal-to-Noise Ratio (PSNR) and Improved Signal-to-Noise Ratio (ISNR) compared to some traditional methods.

The E-Bus is a peripheral parallel interface of a self-developed Digital Signal Processor (DSP) chip named YHFT-DSP. It could support various synchronous or asynchronous commerce standards of host-device interface protocols in order to exchange data effectively between them. In order to design, test and apply the E-Bus interface effectively, a set of simulation platform based on E-Bus was implemented. It used hardware/software cooperation and Field Programmable Gate Array (FPGA) technology to carry out the protocols transformation between E-Bus and USB2.0. The results of test show that the platform has a nice operable interface and implements the data exchange functions between host-device and YHFT-DSP under the master or slave modes of E-Bus.

Active queue management is a focus of current research. Random Early Detection (RED) is one kind of classical queue management algorithms. Linear RED is simple and easy to calculate; however, when average queue size is near to the minimum and maximum threshold, the loss rate is unreasonable. After verifying the nonlinear character between average queue size and packet loss rate, an improved RED algorithm named JRED was presented. The simulation on NS2 shows that the average throughput is improved, and the packet loss rate is decreased. With the JRED algorithm, the stableness and reliability of network are enhanced.

Lexical stress is an important prosodic feature, especially for stress-timed language such as English. To overcome the defects of fixed frame-length features, pitch synchronization feature analysis method was proposed while Pitch Synchronization Energy (PSE) and Pitch Synchronization Peak (PSP) features were defined and extracted. Their contributions, along with traditional features and their combinations, to English lexical stress detection were evaluated with ISLE database. Experimental results show that the combination of new feature and traditional features demonstrates a 6.65% error rate reduction compared with using traditional ones.

An automatic multi-view point cloud merging algorithm for structure light measure system was presented. Firstly, an improved region identification integrated with least-squares algorithm was used to compute the center of the reference point accurately, and the 3D coordinate of the reference point was ascertained through the theory of machine vision. Then, based on the invariability of spatial character, a rapid reference point matching algorithm was put forward to acquire the corresponding relationship between the reference points with different visual angles. Lastly, the rotation matrix and translation matrix were computed by quaternion decompose algorithm to carry out the automatic multi-view point cloud merging algorithm. The result of the experiments shows the validity and great practicability of the algorithm.

Removing duplicated Web pages can improve the searching accuracy and reduce the data storage space. Current deduplication algorithms mainly focus on keywords deduplication or semantic fingerprint deduplication and may cause error when processing Web pages. In this paper each character was mapped into a semantic value by KarhunenLoeve (KL) transform of the relationship matrix, and then each document was transformed into a series of discrete values. By Fourier transform of the series each Web page was expressed as several Fourier coefficients, and then the similarity between two Web pages was calculated based on the Fourier coefficients. Experiment results show that this method can find similar Web pages efficiently.

With network speed and the number of rules constantly increasing, pattern matching is becoming the bottleneck in Network Intrusion Detection System （NIDS）. This paper proposed a fast Wu-Manber-like multi-pattern matching algorithm for intrusion detection, called FWM. By subdividing the pattern group into two subgroups and dealing with the two subgroups in different methods, the FWM algorithm enhanced the efficiency of pattern matching. Experimental results show that, when pattern group contains the pattern that is less than three bytes, the FWM algorithm improves average performance by 29%~44% compared to the original NIDS pattern matching algorithm.

In traditional graph theory based image segmentation methods,the grayscale value of an image is processed directly to obtain clustering results, but the computing time of these methods is very large. A novel segmentation method based on graph partition on histogram clustering was presented. The proposed algorithm obtained threshold by clustering histogram potential function. Since the input is histogram data, the computation time will not be affected by the image size. Experiment results demonstrate that the computation time can be significantly reduced by the proposed algorithm.