摘要
针对海洋渔业遥感信息与资源评估服务系统中中心渔场预报的问题 ,提出了一种多策略的基于CBR的趋势预测方法 ,通过不同的相似性度量方法 ,复用完全相似或者条件相似的历史序列 ,利用领域专家规则对预报结果进一步修正 ,取得了较好的预报结果。相关系统正在应用推广之中。
Focusing on central fishery prediction in system of remote sense information and resources evaluation service in sea, a multi strategies trend prediction method based on CBR is proposed. Adapting many different similarity calculation methods, reusing quite similar or conditional similar historical series, and reviewing prediction results further with rules of domain experts, the results are satisfying. The System is applying to East China Sea fishery area.
出处
《高技术通讯》
EI
CAS
CSCD
2001年第5期64-68,共5页
Chinese High Technology Letters
基金
863计划资助项目 ( 863 818 0 7)
