In this paper, we introduce the stochastic power cover (SPC) problem, which aims to determine the two-stage power assignment and minimize the total expected power consumption. For this problem, we are given a set U of n users, a set S of m sensors on the plane and k possible scenarios, where k is a polynomial and each consists of a probability of occurrence. Each sensor s∈ S can adjust the power it produces by changing its radius and the relationship between them satisfies the following power equation p( s ) =c· r ( s ) ^α . The objective is to identify the radius of each sensor in the first stage and augment the first-stage solution in order to cover all users and minimize the expected power over both stages. Our main result is to present an O( α )-approximation algorithm by using the primal-dual technique.
Bisimulation is a famous behavioral equivalence relation for discrete event systems and has developed rapidly in model checking. Inspired by the bisimulations theory of nondeterministic fuzzy Kripke structures(NFKSs) proposed by Deng [26], in this paper, we define the concept of ϵ -approximate bisimulations for NFKSs. Next, we introduce the notion of the set of traces. On above basis, we propose a new mapping which is used to compare the behaviors of two NFKSs under ϵ -approximate bisimulation. Meanwhile, we investigate the property of finite paths. This property lays the groundwork for exploring the behaviors of two NFKSs under the ϵ -approximate bisimulation. Fortunately, we demonstrate that if two NFKSs equipped with an ϵ -approximate bisimulation between them, then their behaviors are ϵ -approximate each other. Furthermore, we discover that there might not exist the greatest ϵ -approximate bisimulation for a given NFKS. At last, in order to complish approximate minimization operation for a given NFKS under ϵ -approximate bisimulation, we elaborate three algorithms to generate all maximal ϵ -approximate bisimulations. Surprisingly, with help of the maximal ϵ -approximate bisimulations, the quotient NFKSs obtained through the Definition 24 not only have minimum number of states, but also the behavior of quotient NFKS differs by ϵ from the behavior of initial NFKS.
Traceability is the ability to trace the usage of artifacts during the software lifecycle process. Though the benefits of establishing a traceability software system have been widely recognized, it is difficult to be applied well in actual development. In this paper, we propose a new method for traceability preservation which may be used in the practical software development process. A formal model for the traceability of software artifacts is first presented, which consists of variable traceability relations, classification and version number controls. We then present the composition, restriction and refinement operations in the software development process. Next, the preservation of traceability under these three operations is discussed respectively. To demonstrate the effectiveness of our approach, we finally develop a prototype tool named Formalized Software Management System (FSMS).
Against the increasing severity of employment difficulties, job recommendation systems for college students are becoming increasingly important. As it is impossible to refer to the students’ own employment data, relevant research based on the school’s historical employment data provides a reference for students’ employment. We proposes a method based on the school’s historical employment data for job recommendation for college students. The specific ideas and characteristics are as follows: First, preprocess the collected data to generate the user portrait. In the user portrait construction process, we proposes using the AHP-Entropy Weight method to construct a weight vector of ability requirements for different positions, highlighting the focus of students’ abilities in different positions. To improve computational efficiency, first, we uses clustering algorithms to construct different user groups with different characteristics. Then, we calculates the similarity between the students to be recommended and the user groups, followed by the similarity with the users in that group to improve computational efficiency. In particular, we prove that if the number of samples in the database is greater than or equal to 6, our algorithm will have a lower average time complexity than the traditional algorithm. To address the scarcity of employment market data for college students, we collects the real employment data of graduating classes of a major in a certain university to build the HNNU-JOB data set based on students’ employment ability characteristics. Extensive experiments on HNNU-JOB show that the proposed method achieves remarkable performance of recommendation.
Multiplex networks are often used to describe the relationship of different properties between the same group of entities in real complex system, in which nodes represent entities and links in different layers represent connections of different properties between entities. The key to link prediction in multiplex networks lies in (1) making full use of the information provided by each layer of a network; (2) effectively fusing the information provided by each layer of the network together. In this paper, we propose a method based on regression and conditional probability, called MRCP, in which the feature vectors of node pairs are the vectors proposed in our previous work. This method combines intralayer probability and interlayer information to predict missing links in multiplex networks. Firstly, the intralayer probability is calculated by using regression algorithm based on intralayer information. Then the conditional probability of link existence is calculated by using the auxiliary layer information. Finally, both probabilities are combined for link prediction. In order to verify the effectiveness of the method, we conducted experiments on 8 real datasets. The experimental results show that the prediction performance of this method is better than compared methods.
