Study On The China Peasant Consumption Structure Optimization Model

China is a large agricultural nation, therefore well-off problem of farmers is the focus problem of nation. Now the consumption structure of farmers is far from the level of well-off, therefore upgrading the consumption structure of farmers is very important for developing the rural market and expansion of domestic demand. How to upgrading the consumption structure of farmers has becoming a key problem of building well-off society in our nation.This article use the combination method of qualitative (8 consumer indicators' panel data of farmers) and quantitative (farmers' consumption data of nearly 6 years in China Statistical Yearbook), based on the status of consumption structure of farmers, analysis the problem of consumption structure of farmers and the factors impacting the consumption structure., and further proposed the strategies and recommendations of upgrading the consumption structure of farmers.In the first chapter, we introduced current literatures of consumption function at home and abroad, analysis the study of consumption problem in western countries and the status of consumption structure model in our country. The second chapter major reviewed the advantages and disadvantages of in Keynesian Theory of Consumption Function, then introduced the contents of consumption structure and the effects and general trends of consumption structure' rational change. The third chapter major expounded the status of consumption structure of farmers in our country. Chapter IV is the main part of this thesis, introduced empirical analysis of the two algorithms.In the chapter IV, we fist introduced the genetic algorithm based on real-coded.Then we general explained the principle of projection pursuit regression algorithm and the proof of convergence rate. In which Projection Pursuit Model to the numerical optimization method of the high-dimensional data into low-dimensional data space projection, reflecting the data structure in order to find the optimal projection of multivariate data processing methods.The third is the consumption structure of this algorithm optimization model proposed in for the best projection direction, used accelerating genetic algorithm that based on real-coded and combined the scatters, carried out on the best projectors of 8 consumer indicators'panel data (including food, clothing, housing, household equipment and services, health care, transport and communications, cultural and educational entertainment products and services, other goods and services etc.) from 2004 to 2009, and further classified and sorted the data, thus obtained the consumption structure of farmers in our country. Finally, we gave a comprehensive review based on the best indicators of the consumption contribution to the objectives of the comprehensive evaluation of the size and evolution, etc.In the last party, according to the best projection of 8 consumer indicators'panel data, we have the summary and outlook. On this basis, we further pointed out the outlet of upgrading the consumption structure of farmers.