Pattern recognition of quantum information based on patterndistance

Looking upon every encoding state of quantum information systems as a quantum information pattern, A kind of pattern-distance between different patterns as a measurement of comparability of quantum information patterns is defined, and two kinds of recognition algorithms based on pattern-distance for quantum information are proposed. They can respectively recognize quantum information with known objective pattern and unknown objective pattern. In the two algorithms, the phases and occurrence probabilities of different eigenpattems of quantum information are sufficiently considered. Two examples demonstrate the feasibility and effectiveness of the two recognition methods. These algorithms point out a new and important path for applications of quantum information and pattern recognition.

Monitoring of Pinus massoniana spatial pattern changes based on RS and GIS techniques

Our research focused on Pinus massoniana information extracted from remote sensing images based on the knowledge detection and decision tree algorithm and established a spatial pattern model, combining quantitative theoretical ecology with remote sensing （RS） and geometric information system （GIS） techniques. Applying information extraction methods and a spatial pattern model, we studied P. massoniana spatial patterns changes before and after the invasion by pine wood nematode （Bursaphelenchus xylophilus） in Fuyang and Zhoushan counties, Zhejiang Province, east China. The P. massoniana spatial patterns are clustering, whether the invasion happened or not. But the degree of clustering is different. Our results show good agreement with field data. Applying the results, we analyzed the relationship between spatial patterns and the invasion level. Then we drew the elementary conclusion that there are two kinds of patterns for pine wood nematode to spread： continuous and discontinuous diffusion. This approach can help monitor and evaluate the changes in ecological systems.

Feature selection algorithm for text classification based on improved mutual information

In order to solve the poor performance in text classification when using traditional formula of mutual information （MI） , a feature selection algorithm were proposed based on improved mutual information. The improved mutual information algorithm, which is on the basis of traditional improved mutual information methods that enbance the MI value of negative characteristics and feature＇ s frequency, supports the concept of concentration degree and dispersion degree. In accordance with the concept of concentration degree and dispersion degree, formulas which embody concentration degree and dispersion degree were constructed and the improved mutual information was implemented based on these. In this paper, the feature selection algorithm was applied based on improved mutual information to a text classifier based on Biomimetic Pattern Recognition and it was compared with several other feature selection methods. The experimental results showed that the improved mutu- al information feature selection method greatly enhances the performance compared with traditional mutual information feature selection methods and the performance is better than that of information gain. Through the introduction of the concept of concentration degree and dispersion degree, the improved mutual information feature selection method greatly improves the performance of text classification system.

An Adaptive Weighted Differential Game Guidance Law

For intercepting modern high maneuverable targets, a novel adaptive weighted differential game guidance law based on the game theory of mixed strategy is proposed, combining two guidance laws which are derived from the perfect and imperfect in- formation pattern, respectively. The weights vary according to the estimated error of the target＇s acceleration, the guidance law is generated by directly using the estimation of target＇s acceleration when the estimated error is small, and a differential game guidance law with adaptive penalty coefficient is implemented when the estimated error is large. The adaptive penalty coeffi- cients are not constants and they can be adjusted with current target maneuverability. The superior homing performance of the new guidance law is verified by computer simulations.