AI帮你理解科学

AI 生成解读视频

AI抽取解析论文重点内容自动生成视频


pub
生成解读视频

AI 溯源

AI解析本论文相关学术脉络


Master Reading Tree
生成 溯源树

AI 精读

AI抽取本论文的概要总结


微博一下
In this paper we have described the use of a methodology for combining expert elicitation and data for parameterisation of Bayesian networks, an important research topic that has been widely acknowledged in the Bayesian networks field but little developed

Parameterisation and evaluation of a Bayesian network for use in an ecological risk assessment

Environmental Modelling and Software, no. 8 (2007): 1140-1152

引用344|浏览34
EI

摘要

Catchment managers face considerable challenges in managing ecological assets. This task is made difficult by the variable and complex nature of ecological assets, and the considerable uncertainty involved in quantifying how various threats and hazards impact upon them. Bayesian approaches have the potential to address the modelling needs...更多

代码

数据

0
简介
  • Decisions are often based on either expert judgement or on complex quantitative models that consider only a small subset of environmental processes within a complex system.
  • These judgements and models typically focus on environmental processes, rather than ecological values, and there is little effort in understanding and quantifying uncertainties associated with complex systems.
  • Bns offer a pragmatic and scientifically credible approach to modelling complex ecological systems, where substantial uncertainties exist
重点内容
  • In environmental management, decisions are often based on either expert judgement or on complex quantitative models that consider only a small subset of environmental processes within a complex system
  • Available data were split randomly so that 80% of data were used for model training and 20% used for testing
  • Model error rates are generated by withholding data of selected variables, and using the model to predict the outcome of the variable of interest
  • For the native fish Bayesian networks, a semi-formal model walkthrough was conducted with experts
  • In this paper we have described the use of a methodology for combining expert elicitation and data for parameterisation of Bayesian networks, an important research topic that has been widely acknowledged in the Bayesian networks field but little developed
  • Other causal network structures (Borsuk et al, 2004; Ticehurst et al, 2005) have been developed using such information sources, unlike this study, parameter estimates for a variable were obtained from only one source
结果
  • A common method of evaluation for a Bn is to measure predictive accuracy.
  • Model error rates are generated by withholding data of selected variables, and using the model to predict the outcome of the variable of interest.
  • Bn evaluation with experts is important.
  • This can be done via a structured review of the model.
  • The authors recognise the need for more data to improve the robustness of model predictions, and to reduce parameter uncertainties.
  • A model deficiency was the inability to differentiate between the responses of different native fish groups, as different groups can respond differently to environmental conditions, such as flow
结论
  • A suite of evaluative methods was used to investigate the uncertainties and inaccuracies in model structure, relationships and outputs (Coupe and van der Gaag, 2002).
  • This process enabled a more targeted approach to the identification of parameters that needed to be accurately quantified and to recommendations for targeted monitoring and studies to collect further information and data
表格
  • Table1: Site information and location of lowland river sites in the Goulburn Catchment
  • Table2: Methodology used to discretise data input variables, and the states of these variables
  • Table3: To update CPTs of expert elicited nodes with data, two sets of trials were conducted, using the EM algorithm
  • Table4: Sensitivity analysis for posterior network (Eildon), showing calculated entropy
Download tables as Excel
引用论文
  • Bhattacharyya, A., 1943. On a measure of divergence between two statistical populations defined by their probability distributions. Bull. Calcutta Math. Soc. 35, 99e110.
    Google ScholarLocate open access versionFindings
  • Borsuk, M.E., Stow, C.A., Reckhow, K., 2004. A Bayesian network of eutrophication models for synthesis, prediction and uncertainty analysis. Ecol. Model. 173, 219e239.
    Google ScholarLocate open access versionFindings
  • Bromley, J., Jackson, N.A., Clymer, O.J., Giacomello, A.M., Jensen, F.V., 2005. The use of Hugin to develop Bayesian networks as an aid to integrated water resource planning. Environ. Model. Softw. 20, 231e242.
    Google ScholarLocate open access versionFindings
  • Clunie, P., Ryan, T., James, K., Cant, B., 2002. Implications for Rivers from Salinity Hazards: Scoping Study. Arthur Rylah Institute, Department of Natural Resources and Environment, Victoria.
    Google ScholarLocate open access versionFindings
  • Cooke, R.M., 1991. Experts in Uncertainty: Opinion and Subjective Probability in Science. Oxford University Press, New York, 321 pp.
    Google ScholarFindings
  • Coupe, V.M.H., Peek, N., Ottenkamp, J., Habbema, J.D.F., 1999. Using sensitivity analysis for efficient quantification of a belief network. Artif. Intell. Med. 17, 223e247.
    Google ScholarLocate open access versionFindings
  • Coupe, V.M.H., van der Gaag, L.C., 2002. Properties of sensitivity analysis of Bayesian belief networks. Ann. Math. Artif. Intell. 36, 323e356.
    Google ScholarLocate open access versionFindings
  • Das, B., 2000. Representing Uncertainties Using Bayesian Networks. DSTO Electronics and Surveillance Research Laboratory, Salisbury, South Australia.
    Google ScholarFindings
  • Dempster, A., Laird, N., Rubin, D., 1977. Maximum likelihood from incomplete data via the EM algorithm. J. R. Stat. Soc. B 39, 1e38.
    Google ScholarLocate open access versionFindings
  • Dorner, S., Shi, J., Swayne, D. Multi-objective modelling and decision support using a Bayesian network approximation to a non-point source pollution model. Environ. Model. Softw., in press. doi:10.1016/j.envsoft.2005.07.020.
    Locate open access versionFindings
  • Gippel, C.J., Finlayson B.L., 1993. Downstream environmental impacts of regulation of the Goulburn River, Victoria. Hydrology and Water Resources Symposium.
    Google ScholarFindings
  • 31 June-2 July 1993, Newcastle, Australia.
    Google ScholarFindings
  • Gori, M., Tesi, A., 1992. On the problem of local minima in backpropagation. IEEE Trans. Pattern Anal. Mach. Intell. 14, 76e86.
    Google ScholarLocate open access versionFindings
  • Hart, B.T., Burgman, M., Grace, M., Pollino, C., Thomas, C., Webb, J.A., Allison, G.A., Chapman, M., Duivenvoorden, L., Feehan, P., Lund, L., Carey, J., McCrea, A., 2005. Ecological Risk Management Framework for the Irrigation Industry. Land and Water Australia, Canberra (Technical Report).
    Google ScholarFindings
  • Henriksen, H.J., Rasmussen, P., Brandt, G., von Bulow, D., Jensen, F.V., 2007. Public participation modelling using Bayesian networks in management of groundwater contamination. Environ. Model. Softw 22 (8), 1101e1113.
    Google ScholarLocate open access versionFindings
  • Korb, K.B., Nicholson, A.E., 2004. Bayesian Artificial Intelligence. Chapman and Hall/CRC Press, London, 364 pp.
    Google ScholarFindings
  • Koehn, J.D., O’Connor, W.G., 1990. Biological Information for Management of Native Freshwater Fish in Victoria. Parks, Flora and Fauna Division, State Government of Victoria, Victoria.
    Google ScholarFindings
  • Laskey, K.B., Mahoney, S.M., 2000. Network engineering for agile belief network models. IEEE Trans. Know. Data Eng. 12, 487e498.
    Google ScholarLocate open access versionFindings
  • Lauritzen, S.L., Spiegelhalter, D.J., 1990. Local computations with probabilities on graphical structures and their application to expert systems. In: Shafer, G., Pearl, J. (Eds.), Readings in Uncertain Reasoning. Morgan Kaufmann, pp. 415e458.
    Google ScholarLocate open access versionFindings
  • Little, L.R., Kuikka, S., Punt, A.E., Pantus, F., Davies, C.R., Mapstone, B.D., 2004. Information flow among fishing vessels modelled using a Bayesian network. Environ. Model. Softw. 19, 27e34.
    Google ScholarLocate open access versionFindings
  • Martin de Santa Olalla, F.J., Dominguez, A., Ortega, J.F., Artigao, A., Fabeiro, C., 2007. Bayesian networks in planning a large aquifer in Eastern Mancha spain. Environ. Model. Softw 22 (8), 1089e1100.
    Google ScholarLocate open access versionFindings
  • McGuckin, J., 2002. An Investigative Study of the Fish Fauna of the Nagambie Lakes and Chateau Tahbilk Lagoon. Streamline Research Pty. Ltd. for the Nagambie Angling Club, Melbourne, Victoria.
    Google ScholarFindings
  • Morgan, M.G., Henrion, M., 1990. Uncertainty: A Guide to Dealing with Uncertainty in Quantitative Risk and Policy Analysis. Cambridge University Press, Cambridge, UK, 332 pp.
    Google ScholarFindings
  • Nicholson, A.E., Boneh, T., Wilkin, T., Stacey, K., Sonenberg, L., Steinle, V., 2001. A case study in knowledge discovery and elicitation in an intelligent tutoring application. In: Breese, K. (Ed.), UAI01, Seattle, pp. 386e394.
    Google ScholarFindings
  • Norsys, 2005. Netica. <www.norsys.com>. Onisko, A., Druzdel, M.J., Wasyluk, H., 2000. Learning {B}ayesian network parameters from small data sets. Workshop on Bayesian and Causal networks, ECAI-2000.
    Findings
  • Pearl, J., 1988. Probabilistic Reasoning in Intelligent Systems. Morgan Kaufmann, San Mateo, CA. Pearl, J., 1998. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. Morgan Kaufmann, San Mateo, CA. Pollino, C.A., 2004. Ecological Risk Associated with Irrigation in the Goulburn-Broken Catchment - Phase 2 - Adverse Changes to Abundance and Diversity of Native Fish. Water Studies Centre, Monash University.
    Google ScholarLocate open access versionFindings
  • Pollino, C.A., Feehan, P., Grace, M.R., Hart, B.T., 2004. Fish communities and habitat changes in the highly modified Goulburn Catchment, Victoria, Australia. Mar. Freshw. Res. 55, 769e780.
    Google ScholarLocate open access versionFindings
  • Pollino, C.A., Hart, B.T., 2005. Bayesian approaches can help make better sense of ecotoxicological information in risk assessments. Aust. J. Ecotoxicol 11, 57e58.
    Google ScholarLocate open access versionFindings
  • Rieman, B.E., Peterson, J.T., Clayton, J., Howell, P., Thurow, R., Thompson, W., Lee, D.C., 2001. Evaluation of potential effects of federal land management alternatives on trends of salmonids and their habitats in the interior Columbia River basin. For. Ecol. Manag. 153, 43e62.
    Google ScholarLocate open access versionFindings
  • Savage, L.J., 1971. Elicitation of personal probabilities and expectations. J. Am. Stat. Assoc. 66, 783e801.
    Google ScholarLocate open access versionFindings
  • Ticehurst, J.L., Newham, L.H.T., Rissik, D., Letcher, R.A., Jakeman, A.J., 2007. Bayesian network approach for assessing the sustainability of Coastal Lakes in New South Wales, Australia. Environ. Model. Softw. 22 (8), 1129e1139.
    Google ScholarLocate open access versionFindings
  • Ticehurst, J.L., Rissik, D., Letcher, R.A., Newham, L.H.T., Jakeman, A.J., 2005. Development of decision support tools to assess the sustainability of Coastal Lakes. In: Zerger, A., Argent, R.M. (Eds.), MODSIM 2005 International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, (December 2005); Melbourne, Australia, pp. 2414e2420.
    Google ScholarLocate open access versionFindings
  • Varis, O., 1997. Bayesian decision analysis for environmental and resource management. Environ. Model. Softw. 12, 177e185.
    Google ScholarLocate open access versionFindings
  • Varis, O., Fraboulet-Jussila, S., 2002. Water resources development in the Lower Senegal River Basin: conflicting interests, environmental concerns and policy options. Water Res. Dev. 18, 245e260.
    Google ScholarLocate open access versionFindings
  • Woodberry, O., Nicholson, A.E., Korb, K.B., Pollino, C.A., 2004a. A Methodology for Parameterising Bayesian Networks. School of Computer Science and Software Engineering, Monash University (Technical Report).
    Google ScholarLocate open access versionFindings
  • Woodberry, O., Nicholson, A.E., Korb, K.B., Pollino, C.A., 2004b. Parameterising Bayesian networks. In: Webb, G.I., Xinghuo, Y. (Eds.), Lecture Notes in Computer Science. AI 2004: Advances in Artificial Intelligence: 17th Australian Joint Conference on Artificial Intelligence, Cairns, Australia, pp. 1101e1107.
    Google ScholarLocate open access versionFindings
0
您的评分 :

暂无评分

标签
评论
数据免责声明
页面数据均来自互联网公开来源、合作出版商和通过AI技术自动分析结果,我们不对页面数据的有效性、准确性、正确性、可靠性、完整性和及时性做出任何承诺和保证。若有疑问,可以通过电子邮件方式联系我们:report@aminer.cn