【正文】 ,2,1( njmi ???? , Where ???? ???mi jijjmi ijj xxmsxmx 121 )(1,1 . 2) Writing out correlative matrix of sample Let correlative matrix nmijrR ?? )( , where the correlative coefficient is ijmi iiij yymr ???? 111 , and 1, ?? iijiij rrr , so R is symmetrical matrix. 3) Calculating eigenvalue and eigenvector According to eigenfunction 0?? IR ? , n eigenvalues can figure out, and n??? ???? 21 According to homogeneous system of linear equations 0)( ?? LIR ? , the eigenvector corresponding to the eigenvalue can figure out, and for nLLL , 21 ? , ),2,1(,),( 21 njlllL Tnjjjj ， ???? 4) Acquiring principal ponents according to the accumulative contribution rate Firstly, the contribution rates of every principal ponent are figured out as follow: 11?? ??????? ?ni ijjb ?? Then according to the rules of accumulative contribution rate as follow to acquire k principal ponents ),2,1( kjZ j ，?? . %85111 ????????????? ??? ?? ni iki i ?? Where, Tmjjjj zzzZ ),( 21 ?? , Tmiiii yyyY ),( 21 ?? . 5) General evaluation making use of principal ponents Finally, we can use the general evaluation value as follow to make decision. ??? kj jj ZbZ 1 Managing the PCA Using the SPSS and the data which are dealt, the Component Matrix can be calculated as table 6: Table 6. Component Matrix of the metrics Metrics\Components 1 2 3 4 5 Life expectancy Infant mortality rate Neonatal mortality rate Maternal mortality ratio 第 11 頁(yè) 共 26 頁(yè) mortality rate of typical epidemic sufferer typical epidemic morbidity Immunization coverage Contraceptive prevalence rate Density of physicians、 nurses， etc hospital beds per 1000 population Total expenditure on health as %of GDP Government expenditure on health as %of Total expenditure on health Private expenditure on health as %of Total expenditure on health Per capita total expenditure on health From the Component Matrix we can descry that in allusion to the first principal ponent, the contribution ing from life expectancy, infant mortality rate and neonatal mortality are much greater, whose absolute value all exceed . Consequently, the first principal ponent can be interpreted as General health status which contains life expectancy, infant mortality rate and neonatal mortality. Similarly, in allusion to the second principal ponent, the contribution ing from mortality rate of typical epidemic sufferer and typical epidemic morbidity are much greater, thus the second principal ponent can be interpreted as the overall epidemic condition which includes mortality rate of typical epidemic sufferer and typical epidemic morbidity. The third principal ponent can be interpreted as financial policy including total expenditure on health as %of GDP, per capita total expenditure on health, government expenditure on health as %of total expenditure on health and private expenditure on health as %of total expenditure on health. In allusion to the forth principal ponent, there are no metrics whose contributions are great, hence we thought the forth principal ponent can be ignored. Finally, in allusion to the fifth principal ponent, the contribution ing from density of physicians, nurses， etc and hospital beds per 1000 population are much greater, thus the fifth principal ponent can be interpreted as the Resources of health systems, which includes density of physicians, nurses， etc and hospital beds per 1000 population. However, there are still more than one metrics in a certain principal ponent. To determine the most important metric in each principal ponent, another method is needed to analyze each metric further. Further analysis As we all know, if there is significant correlation between variables, any one of them could be used to reflect all the characters of them. On the other hand, in allusion to a certain metric, the bigger the difference of it between each country is, the more information it carries. What’s more, the information taken by a metric can be scaled by Information Entropy, the bigger the Information Entropy is, the more information it carries. Combining both two aspects mention above, the metric which we need is the one who has the biggest Information Entropy between some interrelated metrics. ? The calculation of Information Entropy 第 12 頁(yè) 共 26 頁(yè) Information entropy is one kind of measurement of uncertainty。 東北大學(xué)軟件學(xué)院 畢業(yè)設(shè)計(jì)（論文）外文參考資料 （原文與譯文） 外文 題目： Health care system 中文題目： 衛(wèi)生保健系統(tǒng) 作 者： Ralph Kimball 專 業(yè)： 軟件工程 班 級(jí)： 軟件 班 學(xué) 號(hào)： 姓 名： 2021 年 6 月 10 日 第 1 頁(yè) 共 26 頁(yè) Health care system Nowadays, the health care systems which are provided by nations have improved their residents’ health remarkably. Take the USA for example, the life expectancy at birth has raised to 80 years old, while the infant mortality has declined to 7‰. In other countries of the world, the similar situation also exists widely. Since the health care systems are playing more and more important roles in national health, people often pay much attention to which systems are better and whether the current systems can be improved. However, the aspects of these systems that vary widely between nations (such as how they are funded, the mode which services are delivered, percentage of GDP spent on health care, the personal living habits and so on) are so many that it is hard for us to determine which system is better. Actually, it is impossible and not necessary for us to use the data about all the aspects, thus our tasks are to choose and bine some metrics to evaluate a health care system with sufficient reasons. Then the evaluation method is used between the USA and another country to check up whether it is available. Using the evaluation