Impact of precision of Bayesian network parameters on accuracy of medical diagnostic systems.
Impact of precision of Bayesian network parameters on accuracy of medical diagnostic systems.
复制标题
DOI：
10.1016/j.artmed.2013.01.004
复制
发表时间：
2013-03
影响因子：
7.5
通讯作者：
Druzdzel, Marek J.
中科院分区：
文献类型：
--
作者：
Onisko, Agnieszka;Druzdzel, Marek J.
贝叶斯网络模型在实际应用中最困难的技术任务之一是获取其数值参数。鉴于这一困难，一个紧迫的问题，一个直接影响的知识工程的努力，是这些参数的精度是否是重要的。在本文中，我们解决实验的问题是否基于贝叶斯网络的医疗诊断系统是敏感的精度，其参数。测试网络包括Hepar II，这是一个用于诊断肝脏疾病的大型贝叶斯网络模型，以及其他六个根据欧文机器学习存储库提供的医疗数据集构建的医疗诊断网络。假设原始模型参数是完全准确的，我们通过将它们四舍五入到渐进的courser尺度来系统地降低它们的精度，并检查这种四舍五入对模型精度的影响。我们的主要结果在所有测试网络中一致，即数值参数的不精确性对模型的诊断准确性影响最小，只要我们避免参数之间的零。实验结果表明，只要避免模型参数中的零值，贝叶斯网络模型的诊断精度不会因模型参数精度的降低而受到影响。
One of the hardest technical tasks in employing Bayesian network models in practice is obtaining their numerical parameters. In the light of this difficulty, a pressing question, one that has immediate implications on the knowledge engineering effort, is whether precision of these parameters is important. In this paper, we address experimentally the question whether medical diagnostic systems based on Bayesian networks are sensitive to precision of their parameters. The test networks include Hepar II, a sizeable Bayesian network model for diagnosis of liver disorders and six other medical diagnostic networks constructed from medical data sets available through the Irvine Machine Learning Repository. Assuming that the original model parameters are perfectly accurate, we lower systematically their precision by rounding them to progressively courser scales and check the impact of this rounding on the models' accuracy. Our main result, consistent across all tested networks, is that imprecision in numerical parameters has minimal impact on the diagnostic accuracy of models, as long as we avoid zeroes among parameters. The experiments' results provide evidence that as long as we avoid zeroes among model parameters, diagnostic accuracy of Bayesian network models does not suffer from decreased precision of their parameters.
登录
查看更多内容
影响因子：
14.4
作者：
Leong, TY
通讯作者：
Leong, TY
影响因子：
7.5
作者：
Gal치n, SF;Aguado, F;Mira, J
通讯作者：
Mira, J
影响因子：
1.5
作者：
Heinrich, Lothar;Pukelsheim, Friedrich;Schwingenschloegl, Udo
通讯作者：
Schwingenschloegl, Udo
影响因子：
7.5
作者：
Kurgan, LA;Cios, KJ;Goodenday, LS
通讯作者：
Goodenday, LS
DOI：
10.1016/0010-4809(68)90016-5
发表时间：
1968-01-01
期刊：
COMPUTERS AND BIOMEDICAL RESEARCH
影响因子：
--
作者：
GORRY, GA;BARNETT, GO
通讯作者：
BARNETT, GO