面向医学诊断的智能决策新方法研究

项目名称： 面向医学诊断的智能决策新方法研究

项目编号： No.61303113

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 陈慧灵

作者单位： 温州大学

项目金额： 23万元

中文摘要： 随着医疗信息化在实际生活中的推广,仅仅依靠医学专家进行诊断已远远不能满足现代社会发展的需求。发挥计算机推理和学习等方面的能力，开发出高效、高准确率的医学诊断智能决策系统辅助医生诊断己成为目前该领域的一个重要方向。本项目拟提出面向医学诊断的智能决策理论模型；在此基础上，结合稀疏学习、主动学习和半监督学习等思想设计快速、高效的医学诊断智能决策方法；基于元启发算法和选择性集成思想研究具有可理解性强和决策精度高的医学诊断智能决策方法。该研究不仅对智能决策方法的进一步发展做出贡献，具有重要的理论意义；而且也为解决以医学诊断为代表的实际复杂决策问题提供了一套科学、有效的解决方案，具有重要的应用价值。

中文关键词： 医学诊断；机器学习；智能决策；支持向量机；极限学习机

英文摘要： With the promotion of medical information in real life, only relying on medical experts for diagnosis has been far from satisfying the needs of the development of modern society. Play computer reasoning and learning ability, developing highly efficient, accurate, intelligent decision-making medical diagnosis systems to help diagnose for doctor has become an important direction of the field. The project is going to propose an intelligent decision theory model for medical diagnosis. Based on this model, we not only designed fast and efficiently intelligent decision-making methods for medical diagnosis by combing with sparse learning, active learning and semi-supervised learning, but also proposed intelligent decision-making methods for medical diagnosis with a strong comprehensibility and accuracy based on the methods of meta-heuristic and selection ensemble. The study not only contributes to the further development of the methods for intelligent decision-making, which has important theoretical significance; but also provides a scientific and effective solution to solve actual medical diagnosis of complex decision problems, which has important application value.

英文关键词： medical diagnosis；machine learning；intelligent decision-making；support vector machine；extreme learning machine