摘要本文致力于对风险投资中的贝叶斯风险决策模型的研究。首先，本文基于贝叶斯风险决策模型，主观经验和风险投资顾问的预测准确度，通过计算，发现最优投资方案为聘请顾问，在顾问认为有利的情况下追加投资1000万，承担的风险为110万。然后，结合区间数定理、排序方法与贝叶斯决策理论，笔者发现在市场对该投资项目需求好的情况下，追加1000万元投资，此时期望收益最高。最后，对贝叶斯风险决策模型进行灵敏度分析，可知决策结果对先验概率的灵敏度较高，对条件概率灵敏度较低。本文发现，贝叶斯风险决策模型有优化、贴近生活、符合经验判断等优点，同时，也有准确性不高、数据处理复杂的缺点。50606

This paper is dedicated to study of Bayesian decision stochastic model in venture capital.

Firstly, based on Bayes Decision Models, by comparing the accuracy of prediction between subjective experience and risk investment consultant, the optimal investment plans are retaining a consultant, adding 1 million to investments when consultant considers it is beneficial and the risk is 1.1 million. Secondly, combining interval number theorem ranking method and Bayes Decision Models, adding 1 million to investments when market demands are advantageous to the investments gives the highest expected revenue. Finally, the sensitivity analysis of Bias risk decision-making model is carried out. The sensitivity of the decision result to the prior probability is higher than that to conditional probability. The paper find out that Bayes risk decision-making model has the advantages of optimization, close to life, in accordance with empirical judgment and so on. However, there are also some disadvantages such as not high accuracy and complex data processing.

毕业论文关键词：风险投资； 贝叶斯决策； 区间数

Keyword: Venture Capital； Bayesian Decision Theory；Interval Number

目    录

1.引言 4

2.风险投资的介绍及假设 5

2.1风险投资介绍 5

2.2假设 5

3.风险决策模型 5

3.1贝叶斯风险决策模型 6

3.1.1风险决策中常用的原则 6

3.1.2贝叶斯决策的理论依据 6

3.1.3贝叶斯决策模型建立 6

3.2区间型贝叶斯风险决策模型及其应用 8

3.2.1区间数 8

3.2.2区间数排序方法 8

3.2.3基于区间数排序方法的贝叶斯风险决策模型 9

3.2.4区间型贝叶斯决策模型建立 9

3.3结果与分析 10

3.3.1基于具体数值的贝叶斯决策 10

3.3.2基于区间数的贝叶斯决策 12

3.4灵敏度分析与误差分析 13

3.4.1先验概率对结果的影响 13

3.4.2条件概率对结果的影响 13

4.优缺点分析及改进方向 14

4.1优点 14

4.2缺点 15

4.3改进方向 15

5.结束语 15

6.致谢