Yuan-Hsin Chen

A novel intrusion detection system based on hierarchical clustering and support vector machines

This study proposed an SVM-based intrusion detection system, which combines a hierarchical clustering algorithm, a simple feature selection procedure, and the SVM technique. The hierarchical clustering algorithm provided the SVM with fewer, abstracted, and higher-qualified training instances that are derived from the KDD Cup 1999 training set. It was able to greatly shorten the training time, but also improve the performance of resultant SVM. The simple feature selection procedure was applied to eliminate unimportant features from the training set so the obtained SVM model could classify the network traffic data more accurately. The famous KDD Cup 1999 dataset was used to evaluate the proposed system. Compared with other intrusion detection systems that are based on the same dataset, this system showed better performance in the detection of DoS and Probe attacks, and the beset performance in overall accuracy.

Temperature prediction and TAIFEX forecasting based on fuzzy relationships and MTPSO techniques

In this paper, we proposed a modified turbulent particle swarm optimization (named MTPSO) method for the temperature prediction and the Taiwan Futures Exchange (TAIFEX) forecasting, based on the two-factor fuzzy time series and particle swarm optimization. The MTPSO model can be dealt with two main factors easily and accurately, which are the lengths of intervals and the content of forecast rules. The experimental results of the temperature prediction and the TAIFEX forecasting show that the proposed model is better than any existing models and it can get better quality solutions based on the high-order fuzzy time series, respectively.

A novel visual secret sharing scheme for multiple secrets without pixel expansion

The main concept of the original visual secret sharing (VSS) scheme is to encrypt a secret image into n meaningless share images. It cannot leak any information of the shared secret by any combination of the n share images except for all of images. The shared secret image can be revealed by printing the share images on transparencies and stacking the transparencies directly, so that the human visual system can recognize the shared secret image without using any devices. The visual secrets sharing scheme for multiple secrets (called VSSM scheme) is intended to encrypt more than one secret image into the same quantity of share images to increase the encryption capacity compared with the original VSS scheme. However, all presented VSSM schemes utilize a pre-defined pattern book with pixel expansion to encrypt secret images into share images. In general, it leads to at least 2x times pixel expansion on the share images by any one of the VSSM schemes. Thus, the pixel expansion problem becomes more serious for sharing multiple secrets. This is neither a practical nor the best solution for increasing the number of secret sharing images. In this paper, we propose a novel VSSM scheme that can share two binary secret images on two rectangular share images with no pixel expansion. The experimental results show that the proposed approach not only has no pixel expansion, but also has an excellent recovery quality for the secret images. As our best knowledge, this is the first approach that can share multiple visual secret images without pixel expansion.

Image copyright protection with forward error correction

A copyright protection method for digital image with 1/T rate forward error correction (FEC) is proposed in this paper. In this method, the original image is lossless and the watermark is robust to malicious attacks including geometric attacks such as scaling, rotation, cropping, print-photocopy-scan, and scaling-cropping attacks and nongeometric attacks such as low-pass filtering, sharpening, JPEG compression attacks. The watermark logo is fused with noise bits to improve the security, and later XORed with the feature value of the image by 1/T rate FEC. During extraction, the watermark bits are determined by majority voting, and the extraction procedure needs neither the original image nor the watermark logo. Experimental results show that not only the image is lossless but also the proposed method can effectively resist the common malicious attacks. Since the proposed method is based on spatial domain and there is no need to do frequency transform, the embedding and extraction performances are quite improved.

An Improved Score Level Fusion in Multimodal Biometric Systems

In a multimodal biometric system, the effective fusion method is necessary for combining information from various single modality systems. In this paper we examined the performance of sum rule-based score level fusion and Support Vector Machines (SVM)-based score level fusion. Three biometric characteristics were considered in this study: fingerprint, face, and finger vein. We also proposed a new robust normalization scheme (Reduction of High-scores Effect normalization) which is derived from min-max normalization scheme. Experiments on four different multimodal databases suggest that integrating the proposed scheme in sum rule-based fusion and SVM-based fusion leads to consistently high accuracy.

A hybrid swarm intelligence algorithm for the travelling salesman problem

: We present a hybrid model named HRKPG that combines the random-key search method and an individual enhancement scheme to thoroughly exploit the global search ability of particle swarm optimization. With a genetic algorithm, we can expand the area of exploration of individuals in the solution space. With the individual enhancement scheme, we can enhance the particle swarm optimization and the genetic algorithm for the travelling salesman problem. The objective of the travelling salesman problem is to find the shortest route that starts from a city, visits every city once, and finally comes back to the start city. With the random-key search method, we can search the ability of the particle and chromosome. On the basis of the proposed hybrid scheme of HRKPG, we can improve solution quality quite a lot. Our experimental results show that the HRKPG model outperforms the particle swarm optimization and genetic algorithm in solution quality.

Detecting pop‐up advertisement browser windows using support vector machines

Abstract New browser windows automatically appear without clicking them when certain web sites are browsed by a user. The user does not desire to open such browser windows but is forced to look at them. Such automatic displays of browser windows sometimes contain important announcements, system notices, web guidance, other related enclosure description, pop‐up advertisements and so on. Moreover, pop‐up advertising browser windows are efficient tools for advertisers. However, huge usage of such advertising has become excessive and caused annoyance for users. This study is to analyze and characterize whole web page syntax to create a trained model from Support Vector Machines to effectively discriminate between pop‐up advertisement browser windows and desirable browser windows. The experimental results show that the overall accuracy of the proposed pop‐up detector is up to 92.11%. Hence, it can really reduce annoyance for users.