Applied Mechanics and Materials Vols. 536-537

Abstract: Multi-label classification (MLC) is a machine learning task aiming to predict multiple labels for a given instance. The widely known binary relevance (BR) learns one classifier for each label without considering the correlation among labels. In this paper, an improved binary relevance algorithm (IBRAM) is proposed. This algorithm is derived form binary relevance method. It sets two layers to decompose the multi-label classification problem into L independent binary classification problems respectively. In the first layer, binary classifier is built one for each label. In the second layer, the label information from the first layer is fully used to help to generate final predicting by consider the correlation among labels. Experiments on benchmark datasets validate the effectiveness of proposed approach against other well-established methods.

Abstract: In the context of globalization, it should be have more advantages if ports join in competition in the form of group. But the port classification is not clear, so the necessary of analyzing the ports’s classification in the Bohai ring area is carried out. Cluster analysis was applied in classifying the ten ports of Bohai ring area and provided the basis for the classification and port layout . According to the clustering result, 10 ports are divided into four categories.

Abstract: In order to improve the accuracy of segmentation, analysis of backtracking-forward maximum matching algorithm exists two defects when dealing with crossing ambiguity, and on this basis, an improved-backtracking forward algorithm for maximum matching algorithm is presented. The improved algorithm is based on the backtracking-forward maximum matching algorithm and adds a module, a chain length of one and 3-words, that can detect and process crossing ambiguity, and taking advantage of counting method, we can merely sort out the defragmenter fields that occurred crossing ambiguity. A number of selected language corpus tests prove that under the premise of the segmentation speed, the improved algorithm can enhance the segmentation accuracy.

Abstract: Conflict among knowledge is commonly seen during fuzzing reasoning. By far, some effective methods have been available for eliminating the conflict among knowledge, they are referred to as conflict resolution methods or conflict resolution strategies. The methods of sorting method according to average with weight and sorting in generalizing sequential relation are new methods proposed in recent year. As verified by researches, sometimes these two methods cannot calculate accurate matching degree in fuzzy reasoning. Therefore, this article proposes a new conflict resolution strategy, i.e., sorting method for approximate matching. The results show that this method is consistent with Hamming distance which is recognized worldwide, thus proving the effectiveness of this method.

Abstract: Air-combat decision modeling in effectiveness simulation has to be concerned with the important feature of decision making, such as complexity, diversity, flexibility. So Several challenges have to be mastered, including: improving the abstract level of modeling, providing friendly modeling language, validating concept model and generated code (or executive model) automatically. In this paper, domain-specific modeling (DSM) method is applied in air-combat decision simulation modeling to cope with those challenges. A graphical and textual domain-specific modeling language (DSML) of air-combat decision is designed through metamodel based on an open source tool, Generic Modeling Environment (GME). A code generator is developed to implement users decision model based on python script.

Abstract: Current fire alarm system can only send fire alarms but failed to report accurately ignition point. So a fire seat intelligent identification system was designed based on wireless sensor networks. The system collects data from wireless sensors using ZigBee wireless communication technology, and carries on cluster analysis on the data set using the improved fuzzy kernel clustering algorithm, and gets accurate clustering results. Finally the ignition source location information is reported to the fire alarm control center. The experimental results show that, compared with other methods, this system realizes real-time monitoring and automatic identification of fire seats, which has virtues of high sensitivity and accuracy and high-speed data transfer.

Abstract: Closeness degree is a metric similar levels between two fuzzy sets. This paper extends the closeness degree to the filed of intuitionistic fuzzy multiple attribute decision making. First it defines the closeness degree of two intuitionistic fuzzy vector, then proposes a method based on closeness degree to solve intuitionistic fuzzy multi-attribute decision making problems with the known weights, and demonstrates the effectiveness of this method with examples at last.

Abstract: The Cognitive Radio (CR) technology is an efficient solution to spectrum scarcity by share the spectrum with the secondary users on a non-interfering basis. The spectrum prediction can rationalize the spectrum allocation based on previous information about the spectrum evolution in time. Against previous spectrum prediction algorithm lack of timeliness and accuracy, this paper proposes a novel approach for spectrum prediction based on Optimally Pruned Extreme Learning Machine (OP-ELM) which improved the original Extreme Learning Machine (ELM) algorithm. This method not only takes the advantage of the ELM extremely fast speed and good precision, but also more robust and generic with additional steps compared with ELM. In order to compare its comprehensive properties to other algorithms, some experiments were designed. The results show that the predictive performance of this new algorithm is more satisfaction than others in spectrum prediction problem.

Abstract: To improve the technical level of grain storage, the knowledge and experience of grain storage，computer and artificial intelligence technology are used to research and develop the early-warning expert system of grain storage security. Furthermore，database，artificial intelligence，management information and decision support system as well as computer and information integration technology are applied to grain storage field. In general, this system is contributed to the intelligence and automation of grain storage management. What’s more, it can maximum be used to reduce the loss rate of the grain in storage process to ensure food security of the country.

Abstract: Aiming at the paradox of D-S evidence theory and computations exponential growth in dealing with large scale conflict evidence combination, a new weighted evidence combination method was proposed, which used conflict coefficient and evidence distance in order to measure the conflict. Through the analysis of single conflict representations weaknesses, compound conflict coefficient has been put forward, meanwhile, the evidence center and current center distance were defined, evidence weight was determined with current center distance and conflict coefficient. The experiment results show that the algorithm settles the paradox effectively, at the same time, computing speed has been greatly enhanced.