Pharmacy Selection And Planning Of City Emergency Medicine Based On Big Data

The development of human society is always fighting against the disease inevitably.In addition to cancer, AIDS and other medical problem of human are facing with all the time, in recent years, many new kinds of diseases and infectious diseases appeared,in China, the 2003 ”SARS”, the 2009”influenza A(H5N1)”,the2013”avian influenza(H7N9)”and many other disease appeared unscrupulously, in foreign countries, in 2014,the Ebola virus in West Africa lead to 7373 people were killed, in 2015,The MERS outbreaked in South Korea resulted in 166.93 people have been isolated and in Russia, the number of influenza deaths is more than 20000 each year. Virus, as a kind of species, has been existing for a much longer time then humans, and brought the most threats in survival, which also lead people to reflect.Modern medical scientists are also working on the measure to prevent and control the infectious disease, in which the distribution of anti infectious drugs is a very effective measure. According to statistics, in China, there are 15 cities whose populations are more then ten million, in such a big city, once a kind of infectious diseases outbreaks,disease transmission rates is unthinkable, the prevention is much more better then treatment, at this point, the effective distribution of drugs that can prevent the infectious is particularly important in the city. Since at the beginning of the distribution, the amount of the drugs is very limited, then how to select some specific pharmacies reasonably under the limited drugs, to get the least amount of distributed pharmacies、the largest amount of beneficiaries and to most effectively reduce the possibility of large-scale outbreak, this is exactly the problem that this paper wants to discuss.The paper will first build the mathematical model according to the problem of pharmacy selection of emergency medicine, and achieve the goal that simplify the indicator set through Lasso regression, further deduce a objective function related with human traffic and distance. Then take Shanghai as an example, select the neededdata using Hadoop big data platform and crawler technology. Using ADMM(Alternating Direction Method of Multipliers)algorithm to decompose the target function into ADMM forms and solve, using ADMM features to realize multicore computing, and propose the evaluation criteria of the result. Finally, we will use the web form to dynamically visualize the data results, achieve the fusion of many kinds of scientific and visualization methods for complex problems.