CAJ | 학술논문

结合印度洋大眼金枪鱼渔业应用背景,提出一种新的面向渔情预测的智能数据处理模型.首先利用支持向量机(SVM)对样本数据进行训练,并运用模糊分类器对支持向量进行规则提取;然后对条件和结论分别进行可拓变换和传导变换,获得渔场动态知识.实验结果表明,获得的动态知识具有更强的解释性,对丰富渔情预测方法、解释渔场形成机制具有重要意义.
결합인도양대안금창어어업응용배경,제출일충신적면향어정예측적지능수거처리모형.수선이용지지향량궤(SVM)대양본수거진행훈련,병운용모호분류기대지지향량진행규칙제취;연후대조건화결론분별진행가탁변환화전도변환,획득어장동태지식.실험결과표명,획득적동태지식구유경강적해석성,대봉부어정예측방법、해석어장형성궤제구유중요의의.
This paper proposed a novel intelligent data processing model which applied to fishery forecasting based on the Indian Ocean bigeye tuna fishery application background. Firstly, obtained the support vectors by training the SVM with the sample data. Secondly, extracted the rules from the support vectors by the fuzzy classifier method. Lastly obtained the fishery dynamic knowledge by the extension transformation for the conditions and the conduction transformation for the conclusions. Experimental results show that the dynamic knowledge is of a forceful interpretive capacity, and it enriches fisheries forecasting methods and reveals the formation mechanism of fishing grounds.