A Study On The Factors Affecting The Selection Of Rural Labor Force Training In Yunnan Province

Since Reform and Opening-up, the government has implemented a targeted and systematic approach to poverty alleviation to solve different forms of poverty and our country has achieved the world’s attention in anti-poverty. Today is the actual stage of anti-poverty. Blood-making poverty alleviation, as a basic method, not only has the characteristics of high poverty alleviation rate and great potential, but also conforms to the idea of sustainable development in China. Rural labor training, as an important way of blood-making poverty alleviation, is also fully implemented by the government. However, the rate of participating rural labor training is low. Thus, it is very important to find the influencing factors of rural labor training options and to put forward some suggestions for improving the rate of participating rural labor training.Under this background, many scholars have conducted the research on the problems of rural labor training. In the study of rural labor training options, scholars used questionnaires to collect the data of Rural Labors and analyzed it with Logistic Regression Method. To consider the influence of family factors and village characteristics on rural labor training options, this paper takes the survey data of rural households in Yunnan Province as the research object. Based on Logistic Regression Method, this paper studies on the influencing factors of rural labor training options in Yunnan province by using the method of machine learning in order to find out the main factors affecting the rural labor training options.Firstly, this paper expounded the basic situation of peasant household from four aspects of human capital, material capital, financial capital and village characteristics. The descriptive analysis of the situation of rural labor training participation and the transfer of rural labor force clarified the importance of training to rural labors also explained the practical significance of the research on the rural labor training options. Secondly, based on related theory, the factors that may influence the rural labor training options are extracted. Using clustering analysis method to verify the explanatory power of the independent variables. In the empirical part, based on the full sample data of Yunnan Province, Logistic regression Method has been used to select the variables that have significant influence on the rural labor training options in Yunnan province. It analyzed the impact of these variables on rural labor training options amply. Comparing the importance of each independent variable in the decision tree learning, bagging, adaboost and random forest method have been used to find the key factors influencing the rural labor training options in Yunnan Province. In order to study whether there are regional differences in the influence factors of rural labor training options, the samples are divided into national regional and non-minority area and this paper uses adaboost and random forest methods to compare the importance of two regional variables. At last, according to the results of empirical analysis, this paper puts forward the corresponding suggestions.Some conclusions are drawn through the study. First of all, the rate of people who does not attend the rural labor training is 77%, of which 79.2% of the rural labor force who does not attend the training is willing to participate in rural labor training. So, the participation rate of rural household rural labor training has a lot of room for improvement. In the next place, the most important factor influencing the rural labor training options is the family financial capital, the second is the family human capital factor, then, the family material resources capital factor and the village characteristic is the last one. The top five most important variables are: total household expenditure, total family income, per capita actual operating land area, average age of household labor and average years of education rural household labor force. Finally, sub regional perspective, the influence of the total family income in the non-minority area is more than that of the national regional. However, the importance of other variables in the two areas showed little difference are similar.