Feature selection is a technique of selecting a subset of relevant features for building robust learning models. In this paper, we developed a data dimensionality reduction approach using sparsified singular value decomposition (SSVD) technique to identify and remove trivial features before applying any advanced feature selection algorithm. First, we investigated how SSVD can be used to identify and remove nonessential features in order to facilitate feature selection performance. Second, we analyzed the application limitations and computing complexity. Next, a set of experiments were conducted and the empirical results show that applying feature selection techniques on the data of which the nonessential features are removed by the data dimensionality reduction approach generally results in better performance with significantly reduced computing time.

MicroRNAs (miRNAs) are newly discovered endogenous small non-coding RNAs (21-25nt) that target their complementary gene transcripts for degradation or translational repression. The biogenesis of a functional miRNA is largely dependent on the secondary structure of the miRNA precursor (pre-miRNA). Recently, it has been shown that miRNAs are present in the genome as the form of polycistronic transcriptional units in plants and animals. It will be important to design methods to predict such structures for miRNA discovery and its applications in gene silencing. In this paper, we propose a parallel algorithm based on the master-slave architecture to predict the secondary structure from an input sequence. First, the master processor partitions the input sequence into subsequences and distributes them to the slave processors. The slave processors will then predict the secondary structure based on their individual task. Afterward, the slave processors will return their results to the master processor. Finally, the master processor will merge the partial structures from the slave processors into a whole candidate secondary structure. The optimal structure is obtained by sorting the candidate structures according to their scores. Our experimental results indicate that the actual speed-ups match the trend of theoretic values.

Adopting two independent models for functional verification and performance analysis respectively could not guarantee the performance models satisfying the functionality correctness. In this paper, a colored Petri nets (CP-nets) based method is proposed to integrate functional verification and performance analysis for network protocols. Firstly, a CP-nets based function model for the protocol is constructed and validated. Then, performance related temporal constrains are added into above model, and data monitor units are generated together to form a corresponding CP-nets based performance model. Finally, based on such performance model, simulation based performance evaluation is executed. Because such coessential CP-nets models are utilized where every occurrence sequence in the performance model corresponds to an occurrence sequence in the functional model, it is guaranteed that both models satisfy the functionality correctness requirements of that protocol. As a representative, an integrated analysis process of TRDP protocol is presented to illustrate the practical effectiveness of our proposed method.

In recent years, structured P2P communications is being widely used for its features of self-organization as well as good scalability and flexibility. To make sure that every node can participate in such a network whether it is behind a NAT or not, we must solve the NAT traversal problem. However, existing NAT traversal methods all need the support of a centralized server which will destroy the distributive characteristic of structured P2P. In this paper, we propose a distributed NAT traversal mechanism called SMBR (Selective-Message Buddy Relaying) for structured P2P. SMBR has two main advantages. The first one is that it does not need the support of a server and thus can maintain the characteristics of structured P2P. Secondly, SMBR uses different mechanisms for the control messages and data according to their size. For control messages, it uses the method of buddy's relay while for data direct connections can be built with the help of the buddy. Using this mechanism, SMBR can achieve a balance between the traversal time and the buddies' load.

The convergence of IP network over broadcasting transmission systems is a feasible solution towards the next generation wireless broadband multimedia network for large numbers of users and Internet service providers. Taking the characteristics of IP over broadcasting network into account, this paper proposes a novel Dynamic Differentiated Service Management (DDSM) to provide QoS guarantee for delivering IP applications on the IP-Broadcasting Gateway equipment. In the DDSM model, the queue threshold is regulated dynamically according to the level of emergency and priority; the scheduling weight derives from the queue saturation during a regulating period. As shown in the performance evaluation and simulation analysis, DDSM significantly improves the resource utility of the IP over broadcasting system, and reduces the packet drop probability and the packet delay. Thus DDSM achieves the effective differentiated service of the diverse IP applications over broadcasting network.

Circulation is an important delivery method for both natural and synthetic molecules, but microenvironment interactions, regulated by endothelial cells and critical to the molecule's fate, are difficult to interpret using traditional approaches. In this work, we analyzed and predicted growth factor capture under flow using computer modeling and a three-dimensional experimental approach that includes pertinent circulation characteristics such as pulsatile flow, competing binding interactions, and limited bioavailability. An understanding of the controlling features of this process was desired. The experimental module consisted of a bioreactor with synthetic endothelial-lined hollow fibers under flow. The physical design of the system was incorporated into the model parameters. The heparin-binding growth factor fibroblast growth factor-2 (FGF-2) was used for both the experiments and simulations. Our computational model was composed of three parts: (1) media flow equations, (2) mass transport equations and (3) cell surface reaction equations. The model is based on the flow and reactions within a single hollow fiber and was scaled linearly by the total number of fibers for comparison with experimental results. Our model predicted, and experiments confirmed, that removal of heparan sulfate (HS) from the system would result in a dramatic loss of binding by heparin-binding proteins, but not by proteins that do not bind heparin. The model further predicted a significant loss of bound protein at flow rates only slightly higher than average capillary flow rates, corroborated experimentally, suggesting that the probability of capture in a single pass at high flow rates is extremely low. Several other key parameters were investigated with the coupling between receptors and proteoglycans shown to have a critical impact on successful capture. The combined system offers opportunities to examine circulation capture in a straightforward quantitative manner that should prove advantageous for biologicals or drug delivery investigations.

Reliable shape modeling and clustering of white matter fiber tracts is essential for clinical and anatomical studies that use diffusion tensor imaging (DTI) tractography techniques. In this work we present a novel scheme to model the shape of white matter fiber tracts reconstructed from DTI and cluster them into bundles using Fourier descriptors. We characterize a tract's shape by using Fourier descriptors which are effective in capturing shape properties of fiber tracts. Fourier descriptors derived from different shape signatures are analyzed. Clustering is then performed on these multi-dimensional features in conjunction with mass centers using a k-means like threshold based approach. The advantage of this method lies in the fact that Fourier descriptors achieve spatial independent representation and normalization of white matter fiber tracts which makes it useful for tract comparison across subjects. It also eliminates the need to find matching correspondences between two randomly organized tracts from whole brain tracking. Several issues related to tract shape representation and normalization are also discussed. Real DTI datasets are used to test this technique. Experiment results show that this technique can effectively separate multiple fascicles into plausible bundles.

Random rotation is one of the common perturbation approaches for privacy preserving data classification, in which the data matrix is multiplied by a random rotation matrix before publishing in order to preserve data privacy. One distinct advantage of this approach is that it can maintain the geometric properties of the data matrix, so several categories of classifiers that are based on the geometric properties of the data can achieve similar accuracy on the transformed data as that on the original data. In this paper, we generalize this idea to the situation where the data matrix is assumed to be vertically partitioned into several sub-matrices and held by different owners. Each data holder can choose a rotation matrix randomly and independently to perturb their individual data. Then they all send the transformed data to a third party, who collects all of them and forms a whole data set for data mining or other analysis purposes. We show that under such a scheme the geometric properties of the data set is also preserved and thus it can maintain the accuracy of many classifiers and clustering techniques applied on the transformed data as on the original data. This method enables us to develop efficient centralized data mining algorithms instead of distributed algorithms to preserve privacy. Experiments on real data sets show that such generalization is effective for vertically partitioned data sets.

With the development of emerging social networks, such as Facebook and MySpace, security and privacy threats arising from social network analysis bring a risk of disclosure of confidential knowledge when the social network data is shared or made public. In addition to the current social network anonymity de-identification techniques, we study a situation, such as in a business transaction network, in which weights are attached to network edges that are considered to be confidential (e.g., transactions). We consider perturbing the weights of some edges to preserve data privacy when the network is published, while retaining the shortest path and the approximate cost of the path between some pairs of nodes in the original network. We develop two privacy-preserving strategies for this application. The first strategy is based on a Gaussian randomization multiplication, the second one is a greedy perturbation algorithm based on graph theory. In particular, the second strategy not only yields an approximate length of the shortest path while maintaining the shortest path between selected pairs of nodes, but also maximizes privacy preservation of the original weights. We present experimental results to support our mathematical analysis.

Key management is a significant part of secure wireless communication. In IEEE 802. Hi standard, 4-way handshake protocol is designed to exchange key materials and generate a fresh pairwise key for subsequent data transmissions between the mobile supplicant and the authenticator. Due to several design flaws, original 4-way handshake protocol cannot provide satisfying security and performance. In this study, we adopt formal specification and verification methods to analyze the 4-way handshake protocol. We give its formal models utilizing two kinds of High-level Petri Nets. Based on these formal models, we use two verification methods, model checking and insecure states deduction, to perform an integrated security verification process. The verification results confirm that the 4- way handshake protocol is vulnerable to Denial-of-Service attack during handshake. To repair such vulnerability, we propose an improved key management scheme named enhanced two-way handshake protocol. According to security analysis and performance evaluation, our proposal could provide stronger security capability and cost less computation and communication time.

Data privacy preservation has become one of the major concerns in the design of practical data-mining applications. In this paper, a novel data distortion approach based on structural partition and Sparsified Singular Value Decomposition (SSVD) technique is proposed. Three schemes are designed to balance privacy protection in centralised datasets and mining accuracy. Some metrics are used to evaluate the performance of the proposed new strategies. Data utility of the three proposed schemes is examined by a binary classification based on the support vector machine. Furthermore, we examine three sparsification strategies. The effect of method parameters on data distortion level and utility is also studied experimentally. Our experimental results on synthetic and real datasets indicate that, in comparison with standard data distortion techniques, the proposed schemes are efficient in balancing data distortion level and data utility. They afford a feasible solution with a good promise for mining accuracy and a significant reduction in the computational cost from SVD.

We present experimental results of comparing the modified K-nearest neighbor (MkNN) algorithm with support vector machine (SVM) in the prediction of condition numbers of sparse matrices. Condition number of a matrix is an important measure in numerical analysis and linear algebra. However, the direct computation of the condition number of a matrix is very expensive in terms of CPU and memory cost, and becomes prohibitive for large size matrices. We use data mining techniques to estimate the condition number of a given sparse matrix. In our previous work, we used support vector machine (SVM) to predict the condition numbers. While SVM is considered a state-of- the-art classification/regression algorithm, kNN is usually used for collaborative filtering tasks. Since prediction can also be interpreted as a classification/regression task, virtually any supervised learning algorithm (such as kNN) can also be applied. Experiments are performed on a publicly available dataset. We conclude that modified kNN (MkNN) performs much better than SVM on this particular dataset.

Maintaining data mining accuracy on distorted datasets is an important issue in privacy preserving data mining. Using matrix approximation, we propose several efficient and flexible techniques to address this issue, and utilize unique characteristics of matrix factorization to maintain data pattern. We use the support vector machine classification to compare accuracy maintenance after data distortion by different methods. With better performance than some classical data perturbation approaches, nonnegative matrix factorization and singular value decomposition are considered to be promising techniques for privacy preserving data mining. Experimental results demonstrate that mining accuracy on the distorted data used these methods is almost as good as that on the original data, with added property of privacy preservation. It indicates that the matrix factorization-based data distortion schemes perturb only confidential attributes to meet privacy requirements while preserving general data pattern for knowledge extraction.

Condition number of a matrix is an important measure in numerical analysis and linear algebra. The general approach to obtaining it is through direct computation or estimation. The time and memory cost of such approaches are very high, especially for large size matrices. We propose a totally different approach to estimating the condition number of a sparse matrix. That is, after computing the features of a matrix, we use support vector regression (SVR) to predict its condition number. We also use feature selection strategies to further reduce the response time and improve accuracy. We design a feature selection criterion which combines the weights from SVR with the weights from comparison of matrices with their preconditioned counterparts. Our preliminary experiments show that the results are encouraging.

A novel approach to noninvasively tracing brain white matter fiber tracts is presented using diffusion tensor magnetic resonance imaging (DT-MRI). This technique is based on successive anisotropic diffusion simulations over the human brain, which are utilized to construct three dimensional diffusion fronts. The fiber pathways are determined by evaluating the distance and orientation from the fronts to their corresponding diffusion seeds. Synthetic and real DT-MRI data are employed to demonstrate the tracking scheme. It is shown that the synthetic tracts are accurately replicated, and several major white matter fiber pathways can be reproduced noninvasively, with the tract branching being allowed. Since simulating the diffusion process, which is truly a physical phenomenon reflecting the underlying architecture of cerebral tissues, makes full use of the diffusion tensor data, including both the magnitude and orientation information, the proposed approach is expected to enhance robustness and reliability in white matter fiber reconstruction.

We investigate the use of the multistep successive preconditioning strategies (MSP) to construct a class of parallel multilevel sparse approximate inverse (SAI) preconditioners. We do not use independent set ordering, but a diagonal dominance based matrix permutation to build a multilevel structure. The purpose of introducing multilevel structure into SAI is to enhance the robustness of SAI for solving difficult problems. Forward and backward preconditioning iteration and two Schur complement preconditioning strategies are proposed to improve the performance and to reduce the storage cost of the multilevel preconditioners. One version of the parallel multilevel SAI preconditioner based on the MSP strategy is implemented. Numerical experiments for solving a few sparse matrices on a distributed memory parallel computer are reported.