A new data mining algorithm is proposed by combining the information gain analysis method with the rough set theory. The attained data of tuna purse seiner output from January 1990 to July 2001 and eighteen environmental factors associated with tuna purse seiner output during the same period and in the same area are analyzed with the proposed algorithm, and a set of key influencing factors affecting the output has been found. The fishery prediction model is established by multiple regression analysis based on the key factors set, which is established by the proposed data mining method. The good prediction performance of the model proves that the key factors setcontainsmaininformationofthoseinfluencingfactorsdata,andthenewdataminingmethodiseffective.% 将信息增益引进粗糙集属性化简算法中,提出一种新的基于粗糙集及信息增益的数据挖掘预测算法。以1990年1月到2001年7月的金枪鱼围网产量及18个影响因子为例进行分析,确定关键影响因子集,建立预测模型检验关键因子集的有效性。实验结果表明,关键因子集包含了影响渔情产量的主要信息,预测模型效果较好,验证了所用预测算法的有效性
展开▼