Towards reliable and explainable models for anticipating ecological change

建立可靠且可解释的预测生态变化模型

• 批准号: DE210101439
• 负责人: Dr Nicholas Clark
• 金额: $ 31.31万
• 依托单位: The University of Queensland
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Early Career Researcher Award
• 财政年份: 2021
• 资助国家: 澳大利亚
• 起止时间: 2021-02-01 至 2024-01-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DE210101439
• 关键词: Towards; reliable; explainable; models; anticipating

This project aims to develop a quantitative framework for multivariate ecological prediction. This will allow us to better anticipate how ecosystems respond to environmental change. Recent modelling advances now make it possible to use the complexity of community ecology data to deliver better predictions. The project intends to use long-term ecological datasets to build and test novel multivariate prediction models, using tick paralysis rates in Australian dogs as a case study. Expected outcomes are better tools for studying ecosystem change and new hypotheses about how ecological communities are shaped. Application of these models should provide significant benefits, such as prediction of paralysis tick burdens to improve risk mitigation.

本计画旨在发展多元生态预测之量化架构。这将使我们能够更好地预测生态系统如何应对环境变化。最近的建模进展现在可以使用社区生态数据的复杂性来提供更好的预测。该项目打算使用长期的生态数据集来建立和测试新的多变量预测模型，以澳大利亚狗的蜱麻痹率为案例研究。预期成果是研究生态系统变化的更好工具，也是关于生态群落如何形成的新假设。这些模型的应用应提供显着的好处，如预测瘫痪蜱负担，以改善风险缓解。