Research Of Staff Promotion And Job-Matching In Enterprise Base On Intetellgence Theory

Promotion is one of the important means of stimulation, but also is the important question studied in human resource management. Employee's enthusiasm for work is impacted by promotion. Through promotion employee gains advance of earning, duty, position, and so on. Opportunities are taken by promotion to advance to a higher position. The promotion inside the enterprise is influenced by gender, age, seniority, education, job, performance, job-matching, and so on. Promotion models are different in corporations. Promotion inside the enterprise is divided into selection of promotion and job-matching and intelligent theory is applied to study promotion question inside the corporation.Many factors of promotion are quantities in this paper, and those factors are input variables, employee's rank in the corporation is output variable. The model of selection is designed by fuzzy neural network. The input variables are divided into fuzzy and non-fuzzy variables, X and Z. The improved Takagi-Sugen model is built to obtain output Y that is the reference value to promotion decision-making. The studying algorithm of extended pi-sigma Artificial Neural Network is suggested to ascertain precondition parameter and conclusion parameter.After choice, the employee promoted is assigned to the right jobs in order to make exertions. Normative detecting table of job qualification is applied to explain the demand of jobs to employee in the different occupations that are education training ability interest character and so on. Matching degree between job and employee is obtained through estimating employee by factors of job qualification. If matching matrix is given and two matrices that mean relation among jobs and among employee are taken into account, the genetic algorithm is applied to solve problem of optimization of job matching. Based on the principle of biological evolution, the replaced rate by comparing between son and father generation is defined and studied qualitatively, and value recommended is given. POX is suggested. Operating of duplicating, operating of crossover, operating of mutation, certain searching strategy and control parameter are devised to gain the job-matching optimization.Based on the method above, demonstration is done by datum of Tianan Insurance Limited Company. It take samples from headquarters of Tianan Insurance Limited Company on eight influent factors of promotion that summed by fuzzy and non-fuzzy factors. Sum of effective samples is 143, 110 samples is as training samples randomly, and promotion model of fuzzy neural network is built by training 70024 times. Right times is 25, and Right rate is 75.76% by 33 samples to prove. It is proved that the relation influence factors of promotion and job ranks is gained exactly by fuzzy neural network model. At same time, using the genetic algorithm designed in this paper, job-matching optimization is gained by program. Comparing with anneal neural network, genetic algorithm in this paper have robustness and capability of searching optimization.