This study examines the discrepancy between big data talent training and industry demand. The study analyzed 85 training programs and over 10,000 job postings from two job boards in China (51job and Zhaopin). Using content analysis, social network analysis, and the BERTopic-TOPSIS model, it mined implicit information from training programs and labeled key competencies in job descriptions. A key finding was a significant supply-demand misalignment: while "data application ability" was a stated goal in 52% of programs, only 11% of graduation requirements specified concrete, measurable skills to achieve it. The study identified three primary employment pathways for big data management and application majors: data management, data analysis, and data platform development. Institutions such as Peking University and Hefei University of Technology were identified as best practices. The study then delineated a cultivation path for the major by integrating the characteristics of these employment pathways, and optimised general knowledge and compulsory courses, core courses, graduation requirements, and the cultivation objectives of the major.

With the increasing requirements for the development and utilization of big data, it becomes more and more important to cultivate big data talents to meet the needs of society. The traditional information literacy talent cultivation program can no longer meet the needs of the society for big data talents. Existing big data management majors focus their curriculum construction on the construction of core courses. They do not pay attention to the correlation between the knowledge points in different core courses. In this paper, we propose a professional knowledge mapping method based on RoBERTa-BiLSTM-CRF model for big data management applications, which can effectively solve the above problems, and can more accurately identify the management relationship between the knowledge points in the core courses, help learners better understand and memorize the knowledge, and then establish the linkage of knowledge and the overall framework.

在优势关系粗糙集方法(DRSA)的框架下,针对不协调的目标信息系统求属性约简.基于优势矩阵的方法是最常用的一类约简方法,但矩阵中不是所有的元素都有效.浓缩优势矩阵只保留对求约简有用的最小属性集,因而可以明显降低约简过程中的计算量.进一步地,浓缩布尔矩阵通过布尔代数的形式有效地弥补了优势矩阵生成效率低的缺点.文中将等价关系上的浓缩布尔矩阵属性约简方法扩展到优势关系上,针对优势矩阵提出了浓缩布尔矩阵的概念,建立了相应的高效约简方法,使效率得到明显提高.最后采用9组UCI数据进行实验,结果验证了所提方法的有效性.

DNA computing is a new method for computation using the technology in molecular biology.The study of DNA computing theory will be of benefit to computing science theory.The series papers systematic discuss the com- putability and the computational capacity of DNA molecular using the formal language and automata theory.In this pa- per,we mainly introduce the grammar structures and the computation methods of DNA splicing model,discuss the computational capacity of several DNA splicing models,and prove the functions which can be calculated by Turning machine will also be work out by DNA splicing model in theory.

Attribute reduction is the most important research topic in rough set theory.The traditional attribute reduction based on discer-nibility matrix can only handle consistent decision tables.Then the concept of improved discernibility matrix was proposed to effectively deal with both consistent and inconsistent decision tables.Further,the condensed Boolean matrix was defined to represent the discernibility matrix in order to save the storage space and improve the efficiency of matrix generation.Based on the previous work,the idea of variable precision was used to select some inconsistent objects in the developing of the discernibility matrix,thus more information can be considered in generating attribute reduction.The experimental results show that the proposed method performs advantages in both running speed and classification accuracy.

It is of great significance to accurately identify the type of disturbance signal to analyze and control the pow-er quality problem.In this paper, a new method of power quality disturbance identification based on matching pursuit optimized by particle swarm optimization ( PSO-MP) and RBF neural network is proposed.Firstly, in order to let the residue signal to better reflect the different disturbance signal difference, the fundamental atomic library is constructed to extract the fundamental frequency signals;Then, the MP algorithm is optimized by PSO to reduce the calculation a-mount, which combines with discrete Gabor atom libraries to accurately extract atomic parameters of residual disturb-ance signal by sparse decomposition, Finally, the RBF neural network is used to identify disturbance signals by fea-tures, which is the mean and standard deviation of the atomic parameter and projection of residual signal on the atom. Simulation examples show that the proposed method can effectively identify several common power quality disturbances with a small amount of computation and good anti-noise performance.

When the theories and methods of set pair analysis are applied to the lattice order decisionmaking, a new decision-making model-set pair analysis of lattice order decision making model is created. Based on the portray of the lattice order of the set pair utility function, the best solution of the decision model is found.

In view of the existence of insufficiency of environmental parameter acquisition control system for the current cold rice seedling house, through researching, develop a sampling control system with low power consumption, high precision, strong versatility, to which MSP430F5438 chip processor with low power consumption is the core. Corresponding with the dynamic range of analog pretreatment system, for the construction of hardware platform, use cold rice seedling period environmental parameters and wave characteristics, the different needs of various stages in seedling period ,to achieve real-time acquisition of the various environmental parameters and intelligent automatic control of ventilation equipment and irrigation facilities, meet the demands of practical application, the system has a good development prospect based on its high reliability and versatility. Keywords-rice seedling-nursery house; MSP430F5438; acquisition controller; general; high precision

Multi-objective decision-making can be divided into two categories: (1) First select some evaluation attributes, then evaluate the pre-programs under these evaluation attributes, and finally using certain methods to fuse the evaluation information; (2) Since the preselected programs in a same external environment, how to select the most pre-match program matching with the external environment. But present research about the later decision-making question is few. This article has conducted some research to this kind of policy-making question As a result of each preselected plan own characteristic, condition-performance functions of various environment condition factor to each preselected plan existence difference. Thus causes not be able to only depend on the value of environment condition factor to judge each preselected plan fit and unfit quality under this environment condition, but must carry on the overall evaluation according to the value of environment condition factor and the condition-performance function. In this article, firstly construct an algorithm to seek consistent fit and unfit quality sorting sector. Then based on these consistent fit and unfit quality sorting sectors, construct a match measurement operator to measure the match degree between each preselected plan and the external environment. The preselected plan owning the biggest match degree is the best plan under this environment condition.

Decision Support System is playing an important role in computer science, technology and engineering. Intelligent decision-making is one of the current hotspots in the decision support system research. Intelligent decision-making methods and algorithms are one of the most important basics and key cores in intelligent information processing, intelligent pervasive computing and so on. In this paper, conduct the research to two kinds of indefinite multi-objective decision making question: the indefinite sector and the indefinite language. (1) In view of multi-attribute decision-making under linguistic setting, propose one new decision method. Firstly construct a range pole plan and introduce the policy-maker risk-preference weight. Then with three tuples (Limit low similarity, Risk degree, Risk-preference value) reflect the risk-degree existing in the decision-making process. At last, construct the risk-weighted similarity measure operator (RWSMO) to measure the risk balance similarity's size between each of decision schemes and the range pole plan. (2) In view of multi-attribute decision-making under the indefinite sector, propose one new decision method based on the multiple-valued intuitive fuzzy sets.

Decision making under uncertainty has very popular application fields. This paper analyzes the difference of decision under risk and under uncertainty, and provides a simple axiom of the Choquet Expected Utility model where the capacity is a subject measure to deal with decision under uncertainty. At last, a reinterpretation of enterprise clustering is given.

There are t hree major difficulties in the indefinite multi-objective decision making process: 1) how to express the indefinite information because of the information about attributes being indefinite ; 2) how to express the indefinite information because of the information has multi-channels ; 3) how to fuse the information into synthetic information. The multiple-valued intuitive fuzzy sets is one new mathematical model, this model can process well fuzzy information gained from multi-sources. In this paper , firstly conduct the research to the multiple-valued intuitive fuzzy set's information fusion and construct some methods to fuse the information included in the degree of membership or non-degree of membership of multiple-valued intuitive fuzzy set , then use the isomorphism mind to research indefinite isomerism multi-objective decision making and construct one new algorithm for interval value and indefinite language isomerism multi-objective decision making based on isomorphism information fusion .

With the development of science and technology, the competition of situation in the market has become more and more fierce. The risk is everywhere. How to reduce risk, avoid risk, minimize risk of the project, is the urgent business, and in the current complicated situation, Cobb-Douglas production function as a mathematical model of the economy, because of its variable value of uncertainty, the economic model can not accurately describe the production of input-output relationship. This paper makes quantitative analysis and scientific choice based on multi-attribute decision-making of the possible degree approach to the feasibility of the path of project in business from the three factors (labor input, capital input and the technical level) which impact the project risk of enterprise, and consequently achieve the enterprise's risk minimization.

Using the parallelization to deal with complex questions,a most basic issue is how to divide the complex question into some sub-questions.Firstly studied some special geometric property about convex shell,then used these property to divide the set of points into some triangle grids.Simultaneously proved convex shell umbo is only located in the periphery triangle grid and looked for convex shell umbo in these periphery triangle region mutually independently.

The cultural algorithm is a high efficiency searching method after evolution algorithm.It rooting from simulation of evolution of human being society,provides a new computable framework for Evolution Algorithms.This paper depicts a improved about cultural algorithm.Compare with primary algorithm,The coordinate was carved up with interzone in the space,and the interzone is incoordinat,then,the semi-feasible region was subdivision into less interzone.So reduce forage scope,increase arithmetic.

There are two major difficulties in the indefinite multi-objective decision making process: As the information about attributes is indefinite, how to express the indefinite information; as the information has multi-channels, how to fuse the information into synthetic information. The multiple-valued intuitive fuzzy sets is one new mathematical model, this model can process well fuzzy information gained from multi-sources. In this article, firstly conduct the research to the multiple-valued intuitive fuzzy set's information fusion and construct some methods to fuse the information included in the degree of membership or non-degree of membership of multiple-valued intuitive fuzzy set , then use the isomorphism mind to research indefinite isomerism multi-objective decision making and construct one new algorithm for interval value and indefinite language isomerism multi-objective decision making based on isomorphism information fusion.

The financial problems often relate to many stochastic differential equations which is difficult to solve. Based on the traditional numerical methods for solving determinate ordinary differential equations (ODEs), the paper presents the numerical methods of stochastic ordinary differential equations (SODEs). The numerical methods mainly considered in the paper include Euler method and Euler method discusses three numerical schemes, i.e. explicit scheme, semi-implicit scheme and implicit scheme. Finally, the paper uses Euler methods to simulate linear stochastic ordinary differential equation. Numerical results indicate that Euler methods are the simple but have a certain practicality. In addition, all these Euler method including explicit, semi-implicit and implicit scheme in this paper can be applied to high-order stochastic ordinary differential equations, stochastic partial differential coefficient equations, and so on.

In recent year parallel computing access to a relatively rapid development, provide a solution. to solve complex problems. With the development of economy, social, science and technologies of which all things are done, seen, studied deeply by human, so the dimension of information has been rising. Thereby creating numbers of large matrixes, there is much computation in processing of the large (including giant) matrix. Parallel computing is an effective solution to solve complex problems, but it involves an important and basic question of how to divide the large matrix into some sub-matrixes or smaller parts of the large matrix. The traditional extant dividing methods (such as band division and chessboard division) have same deficiencies: communication intensity between the processors is high. In order to overcome this weakness and realize large matrix parallel transpose efficiently, based on decreasing the communication intensity among the processors, some isomorphic new parallel dividing methods for special large matrix transpose are studied and given in this paper.