CAJ | 학술논문

C4．5算法是目前应用最为广泛的一种决策树算法，针对其分支策略和属性选取的不足，提出与分裂属性和类属性相关的平衡因子，以修正属性的分裂信息，进而选取更有意义的属性作为分裂节点，协调各属性的信息增益率。采用经典天气实例与原C4．5算法对比，改进的决策树分类更为合理、准确。改进C4．5算法用来预测不同参数下SM T 焊接质量以选择最佳参数，可提高SM T生产效率。
C4．5산법시목전응용최위엄범적일충결책수산법，침대기분지책략화속성선취적불족，제출여분렬속성화류속성상관적평형인자，이수정속성적분렬신식，진이선취경유의의적속성작위분렬절점，협조각속성적신식증익솔。채용경전천기실례여원C4．5산법대비，개진적결책수분류경위합리、준학。개진C4．5산법용래예측불동삼수하SM T 한접질량이선택최가삼수，가제고SM T생산효솔。
C4 .5 algorithm is currently the most widely used decision tree algorithm .In view of the decision tree C4 .5 algorithm branching strategy and attribute selection ,the balance factors associated with division and class attribute is put for-ward .The factors can be applied to modify the division of property information ,to choose a more meaningful attribute as the split node and to coordinate information gain ratio of various properties .The classic example weather can be compared to the improved C4 .5 algorithm and the original C4 .5 algorithm ,the improved decision tree classification is more reasonable and ac-curate .The improved C4 .5 algorithm is used to predict the SMT welding quality under different parameters to choose the best parameters ,can improve the efficiency of the SM T production line .