Nonlinear Hypothesis Generation Strategy of Management Accounting Data Mining

Abstract

Management accounting plays a greater role in providing business decisions for enterprises. As the core technology of big data processing, data mining technology can process a large amount of complex, structured, and unstructured information very efficiently. In this paper, the nonlinear system is taken as the research object, and the obtained system data is directly used. In the framework of divide and conquer strategy, multi-model method is combined with data clustering and a local modeling algorithm to study the nonlinearity of management accounting data mining. Through online adaptive estimation of parameters, aiming at the zero dynamic problem of the non-minimum phase system, through the decoupling matrix analysis of the nonlinear system, the state of the original strategy is dynamically extended to make the estimated parameters asymptotically approach the true value. The estimated parameters are applied to the nonlinear hypothesis strategy to achieve the goal of linearization control. It is assumed that the generation strategy is applied to the mining of different related types of rules without manually setting parameters, which greatly improves the efficiency of rule discovery. Through case analysis, this paper demonstrates the feasibility of applying nonlinear hypothesis strategy of data mining technology to management accounting and puts forward the relevant application process framework. The validity and feasibility of nonlinear hypothesis strategy data mining technology in management accounting are verified.