The Research Of One Multinational Company Staff Turnover Prediction Model Based On COX Model

The impact of the BRIC forum,led by China and Asian infrastructure investment and Development Bank on the world's economic situation is growing. In addition, many going out strategy is put forward, such as the One Belt And One Road, it is not hard to see in the sustained and rapid development of global economic integration, China matters increasingly to the world economy. In order to maintain strong competitiveness in the increasingly fierce competition in the global market, we must optimize the structure of the Chinese enterprises, enhance the management level, improve management mode. And it is necessary to make reformation and Innovation in the existing human resources management Contents of the abstract. Times New Roman.In 2016, Two Sessions governmental work report emphasized that the third industry, headed by services, in our country has more than 50% of GDP accounted. Fast consumer industry is an important part of the service industry. The employee of Fast consumer industry is uneven, the proportion of the market and channel personnel is very big, at the same time, they lack of effective personnel training system. Because of the complexity of its management and the objective conditions limitation, such as the existing domestic management system, the staff turnover rate of Fast away industry is very high. Therefore, the introduction of international management experience to improve its scattered, decentralized management status can not only ensure quick elimination enterprise itself does not decline in the fierce market competition, but also has strategic significance to the steady and rapid development of the whole Chinese economy.The data in this research contains 7650 items which comes from the human resources management system of a certain domestic COLA company. This research accurate extraction, conversion and clean this data, using latest data mining techniques and analysis methods at home and abroad, finally obtained 5703 experimental data. Using R software, the author compares and analyzes the decision tree, logistic regression and the semi parametric model of survival analysis, finally choose the COX model to structure personnel loss prediction model. The experimental data obtained multi angle evaluation and inspection from all directions.By the model analysis results, the prediction accuracy of the model is as high as 85%. In addition, the model takes the censored data into account, so the model is more effective. Moreover, the model gives the cumulative probability of survival of employees. Finally, we find out the subjective and objective reasons of the personnel loss of Chinese FMCG enterprises and the key point of the reform to avoid this kind of loss, thus, can provide the data reference for the implementation of talent retention measures.