DNA Sequence Classification Based on the Side Chain Radical Polarity of Amino Acids

The features of DNA sequence fragments were extracted from the distribution density of the condons in the individual cases of DNA sequence fragments. Based on the polarity of side chain radicals of amino acids molecules, the amino acids were classified into five categories, and the frequencies of these five categories were calculated. This kind of feature extraction based on the biological meanings not only took the content of basic groups into consideration, but also considered the marshal ing sequence of the basic groups. The hierarchical clustering analysis and BP neural network were used to classify the DNA sequence fragments. The results showed that the classification results of these two kinds of algo-rithms not only had high accuracy, but also had high consistence, indicating that this kind of feature extraction was superior over the traditional feature extraction which only took the features of basic groups into consideration.

Optimization of satellite searching strategy of the non-stationary antenna

The tiny searching step length and the satellite distribution density are the major factors to influence the efficiency of the satellite finder,so a scientific and reasonable method to calculate the tiny searching step length is proposed to optimize the satellite searching strategy. The pattern clustering and BP neural network are applied to optimize the tiny searching step length. The calculated tiny searching step length is approximately equal to the theoretic value for each satellite. In application,the satellite searching results will be dynamically added to the training samples to re-train the network to improve the generalizability and the precision. Experiments validate that the optimization of the tiny searching step length can avoid the error of locating target satellite and improve the searching efficiency.

Applied Fault Detection and Diagnosis for Industrial Gas Turbine Systems

The paper presents readily implementable approaches for fault detection and diagnosis （FDD） based on measurements from multiple sensor groups, for industrial systems. Specifically, the use of hierarchical clustering （HC） and self-organizing map neural networks （SOMNNs） are shown to provide robust and user-friendly tools for application to industrial gas turbine （IGT） systems. HC fingerprints are found for normal operation, and FDD is achieved by monitoring cluster changes occurring in the resulting dendrograms. Similarly, fingerprints of operational behaviour are also obtained using SOMNN based classification maps （CMs） that are initially determined during normal operation, and FDD is performed by detecting changes in their CMs. The proposed methods are shown to be capable of FDD from a large group of sensors that measure a variety of physical quantities. A key feature of the paper is the development of techniques to accommodate transient system operation, which can often lead to false-alarms being triggered when using traditional techniques if the monitoring algorithms are not first desensitized. Case studies showing the efficacy of the techniques for detecting sensor faults, bearing tilt pad wear and early stage pre-chamber burnout, are included. The presented techniques are now being applied operationally and monitoring IGTs in various regions of the world.

Two-level hierarchical feature learning for image classification

In some image classification tasks, similarities among different categories are different and the samples are usually misclassified as highly similar categories. To distinguish highly similar categories, more specific features are required so that the classifier can improve the classification performance. In this paper, we propose a novel two-level hierarchical feature learning framework based on the deep convolutional neural network(CNN), which is simple and effective. First, the deep feature extractors of different levels are trained using the transfer learning method that fine-tunes the pre-trained deep CNN model toward the new target dataset. Second, the general feature extracted from all the categories and the specific feature extracted from highly similar categories are fused into a feature vector. Then the final feature representation is fed into a linear classifier. Finally, experiments using the Caltech-256, Oxford Flower-102, and Tasmania Coral Point Count(CPC) datasets demonstrate that the expression ability of the deep features resulting from two-level hierarchical feature learning is powerful. Our proposed method effectively increases the classification accuracy in comparison with flat multiple classification methods.