摘要本文是为了确定最佳的投资策略，包括大学的分类、排名、投资回报率计算以及投资持续时间模型。             

首先，本文提出了基于贝叶斯判别方法的大学分类模型，由此把大学分为5类。在此基础上，基于k-均值聚类算法的测试模型显示了前者分类模型的有效性。             

为了确定高校教育财政需求的权重，建立了一个层次分析法和熵值法相结合的模型，对学校进行赋值排名。89396     

针对投资回报率，提出了一种基于逻辑回归的投资回报率计算模型。在分析了相关因素后，根据经济学原理定义了投资回报率的公式并提供了合理的测试模型。

最后，我们建立了基于投资Ito方程投资时间对学生表现影响的预测模型。模型提供了最佳的投资时间持续时间。除此之外，设计了两个捐赠计划。一是将捐赠分为助学金和助学贷款。另一类是投资稳定的理财产品。由此提高资金利用率和增加投资年限。

综上本文给出了针对各个问题的解决方案，但是最主要的缺点是缺乏足够的数据来检验模型的有效性。

In this article, we develop models to determine an optimal investment strategy, including University Classification Model, Ranking Model, ROI Calculation Model and Time Duration Model。

Firstly, a model for university classification using Bayes Discrimination Method is provided based on 7 parameters。 We pide universities into 5 categories and get the most optimal choice。 After that, a testing model based on k-means Clustering Algorithm shows the validity of the classification model。 The model can be used not only in university classification, but also any classification problem。

To determine the weight of university’s demand of educational finance, we build a model, which is the combination of Analytic Hierarchy Process and Entropy。 And we choose NPT4_PUB, PCTPELL, PCTFLOAN, GRAD_DEBT_MDN_SUPP, RPY_3YR_RT_SUPP as correlative factors。 Then we draw conclusions of which universities we should donate and their donations。 

A ROI Calculation Model based on Logistic Regression is put forward。 We analyze the relation among SAT_AVG, gt_25k_p6, C150_4_POOLED_SUPP and PCTPELL。 Then we define the formula of ROI。 Reasonable test of the model is also provided。

Finally, we establish the student performance prediction model based on the duration of the investment ITO equation。 The max derivative of ITO equation is the highest likelihood of producing a strong positive effect on student performance。 And the best investment time duration is also provided。 Apart from this, we design two donating programs。 One is piding donations into grants and student loans。 The other is investing stable financial products。

Sensitivity analysis of our models is provided, and a primary weakness of our work is a lack of enough data to test some of our models。

毕业论文关键词：聚类分析； t检验； 投资回报率： Ito方程

Keyword: Clustering Analysis: T-test: ROI: ITO Equation

目   录

1。引言 4

2。学校分类模型 5

2。1问题分析 5

2。2数据预处理 6

2。3基于贝叶斯的源Q于W优E尔A论S文R网wwW.yOueRw.com 原文+QQ75201,8766 分类模型 7

2。3。1聚类分析 7

2。3。2正态性检验 8

2。3。3贝叶斯判别法 8

2。4基于k-均值聚类算法对比模型检验