| Food security,energy security,and financial security are regarded as the three big economic security in this world,but food security is the most fragile.The state security law,which was through on July 1,2015,put forward explicitly to improve the food security system and put food security as an important part of national security strategy.With the advent of the era of big data,the three-dimension includes cyberspace,physical world,and human society continue to interleave,collaborate,and merge together,the data reflecting the food security is becoming mass,multi-source,and real-time,the intelligent model,calculate platform are becoming more innovative,so those provides a new perspective for food security early warning.This thesis,from the big data visual angles,take China's food security early warning as the research object,on the basis of summarying ditional food security early warning theories,methods and technologies,to build food security early warning theory of rational expectations with the the big data supporting,and then add up to some different attribute dimension of early warning indicators involving food security,Dimensions of different properties,such as commodities,weather,finance,and energy,and extract characteristics analysis.Based on the early warning indicators to build the prediction model of grain supply and demand,and take the maize to carry on the empirical analysis and conducted early warning analysis on price expectations using big data,at last put forward some policy suggestions from the perspective of big data.In this thesis,in the following several aspects to carry out the innovative research:Firstly,introduces the theory of rational expectations on food security early warning.In absorbing expectations theory,information theory,decision science and early warning theory,develop a food security early warning theory of rational expectations,which means big data can eliminate uncertainty,Super calculate system can improve information processing efficiency,expectation management can reduce the information asymmetry,make early warning and decision-making looming more rational.Market expectations can be identified,managed and even be recreated.In the process of the flow of information and communication,who mastered the authority,transparency,and interpretation of information and who have demonstrated the controls the market expectations of the initiative,influence and voice.Secondly,food warning indicators system based on large data has been constructed,which has broken the limitation of past research indicators on food security.Some new influence factors such as climate,capital,energy indexeshave been added into the early warning index system,such as the southern oscillation index,non-commercial holdings of futures markets,biomass energy development situation.On the background of globalization,food security index early warning research under complicated environment,has improved the completeness of food security early warning indicators.Thirdly,Basic research such as threshold and cycle of early warning indicators is carried out.In the aspect of the warning index threshold,using quantile method and the probability distribution method determines the threshold of timely warning.Using the X11 method and spectral density method to do the researches of early warning indicators for the seasonal decomposition and cycle recognition.Discover the food production fluctuation cycle is about 13 years,rice is about 7 years,wheat and corn for 3 years or so.Fourthly,build the prediction model of grain supply and demand based on maize as an example.Based on United Nations high population growth predicted,we make a forecast analysis of the supply and demand of maize in the next decades.The results show that between 2016 and 2020,if the domestic consumption grows in the average annual growth of 4.9%,maize area decline in the average annual of 3.4%,maize yield should keep the average annual growth of 1.6%,and the annual inventory should reduce to 13 million tons,and the import maintain at 1.5 million tons,the domestic maize can realize the balance between supply and demand;In next 10 years,if the domestic consumption increase in the average annual growth of 3.0%,maize area decline in the average annual of 0.8%,maize yield should increasing 1.2% per year,inventory reduces at the speed of annual 10 million tons,import maintains at an annual 2.4 million tons,domestic maize can realize the balance between supply and demand.Fifthly,this thesiscarried out large data application in price expectations cautioning.we took the maize industry chain price information flow as the research object,1)in view of the overall price level to change,build the price average system,identify the change price trend and market expectations,build the regional early warning radar chart and seasonal price as figure,find the key locations and key point of the price trend and changes;2)For abnormal fluctuations persistent identification problem.According to the principle of probability distribution function and 3-sigma principle,persistent identification model constructed price changes and warning rules,and effectively identify the up and down,one-sided,continuous fluctuations and other anomalies;3)for price expectations associated problems.We carried out the correlation analysis about home and abroad maize price difference,spot prices and futures nearly far spread,elaborates the relationship and comparison between the grain varieties linkage effects.Food security system structure of multilateral,process stages,compositional heterogeneity,dynamics and nonlinear determines the food security early warning system is a very complicated and difficult task.Suggestions from the perspectives of big data,through the perfect food industry chain monitoring,strengthening the data intelligent analysis,in order to strengthen the management of information sharing,strengthening the expected to guide,rich food control toolbox,food pre-warning safety net,to promote food security early warning continuously develop toward a real-time,accurate and scientific direction. |
PDF Full Text Request