Haibo Yang

A new trust model for online social networks

Online social media networks play important roles for people to share opinions, communicate with others. One of important features behind these activities is trust. This paper investigates the trust model in Online social media networks. Considering the interaction between two users and the reputation in the social networks, this trust model gives a definition about the trust value between two users. The Gaussian kernel density estimation is used to describe the users reputation among a social network. Combined with the interaction relationship, the trust model can provide more accurate trust relation prediction. Experiments on the public dataset shows the method generates high accuracy results.

Combining Statistical Information and Distance Computation for K-Means Initialization

As the symbol of the partition clustering method, K-Means is well known and widely used in many fields for the easily implemented and high efficiency. However, the initial center problem may affect the final cluster result, sometimes the final cluster result might contain some empty clusters. In this paper, a new K-Mean initialization method is proposed which combines the statistical information and the distance computation. The statistical information contains the mean, median, and Gaussian kernel density estimation. At first, the high density points are selected for each dimension. Then the distance and the density are used to measure every possible initial centers. After this process works from high variance dimension to low variance ones, the final initial cluster centers are constructed with the K nearest neighbors. Experiments on public datasets show that this method can achieve comparable results compared with other conventional methods.

Building Influence for Online Social Networks.

Online social networks have been changing people’s lives from personal profiles, social activities to the method people communicate with others and commercial advertisements. Among all those activities, personal influence is an important factor. To some extent personal influence affects the final results of almost every action. In this paper, a new influence model is proposed combining the global influence and local influence. The global influence is based on the follower relationship, which is computed with Gaussian kernel density estimation. The local influence focuses on one’s influence for the communities he joins. Recognized degree and personal capability in the community are employed for computing the local influence value. The final influence value is computed with a linear combination between the global influence and the local influence. The proposed method is evaluated with a public dataset Flickr. Results show that the proposed method can provide accurate prediction for personal influence.

A new hierarchical method for music genre classification

Large collections of music bring new challenges for people to choose the favorite songs. Music genre explicitly defined can help people to solve this problem. However, classifying the music genre automatically is a challenging problem since many genres do not have any special features. This paper presents a new music genre classification method which utilizes hierarchical analysis of the spectrograms features extracted from the audio signals. First support vector machines are used to build the classification tree, then K nearest neighbors are implemented to improve the accuracy of the classification. The GTZAN genre collection music database is used to evaluate the proposed model, and the results show that this model can get comparable results compared with some other existing music genre classification methods.

Content-based music similarity computation with relevant component analysis

Content-based music similarity is becoming important because of the millions of songs with online distribution. However, current methods have to treat the same attention on unrelated information or different informative information. In this paper, a new method is proposed to compute the music similarity with relevant component analysis. Considering the different weights for different parts, this method pays more attention on those parts with informative features by ranking the Mel-frequency cepstral coefficients frames on energy, and masters the song features more precisely compared with other methods. Experimental results on public dataset MusiClef show that this method works faster on music similarity computation task without sacrificing the accuracy of the similarity measurement.

A new projection-based K-Means initialization algorithm

As a partition based clustering algorithm, K-Means is widely used in many areas for the features of its efficiency and easily understood. However, it is well known that the K-Means algorithm may get suboptimal solutions, depending on the choice of the initial cluster centers. In this paper, we propose a projection-based K-Means initialization algorithm. The proposed algorithm first employ conventional Gaussian kernel density estimation method to find the highly density data areas in one dimension. Then the projection step is to iteratively use density estimation from the lower variance dimensions to the higher variance ones until all the dimensions are computed. Experiments on actual datasets show that our method can get similar results compared with other conventional methods with fewer computation tasks.

Robust control for TCP network systems with input delay

In this paper, the robust control problem is studied based on a previously developed nonlinear dynamic model of TCP. By using Lyapunov-Krasovskii stability theory, a state feedback controller is designed to guarantee that the system could be robust asymptotically stable. Firstly, the packet-dropping probability is considered as an input, by linearizing the interconnection of TCP, an uncertain system with input delay is obtained. And then, by a state transformation, the original input-delayed system is converted into a state-delayed system. For the novel TCP network mode, the Lyapunov-Krasovskii function is defined, and the TCP network system can be asymptotically stabilized with the obtained state feedback control law by using Linear Matrix Inequalities (LMIs) approach. Finally, theory analysis and simulation results show the effectiveness of the derived control strategy and superiorities compared with the existing results.