frequent pattern d 数据仓储与数据挖掘 方案策划.pptVIP

关联规则 Mining various kinds of association rules ;Mining Various Kinds of Association Rules;Multi-level AR mining;多层挖掘算法;使用不同的minisupp，方法1;使用不同的minisupp，方法4;去掉冗余的多层规则;Mining Multi-Dimensional Association;Mining Quantitative Associations;Mining Quantitative Associations;Mining Quantitative Associations;Mining Quantitative Associations; Mining Quantitative Associations;第二步: 扫描二维数组, 寻找频繁谓词集, 产生关联规则 第三步: 借用cluster方法, 合并关联规则 如: age(X, 34) ∧income（X , “3000-4000” ）→buys（X , “color TV” ） age(X, 34) ∧income（X , “4000-4900” ）→buys（X , “color TV” ） age(X, 35) ∧income（X , “3000-4000” ）→buys（X , “color TV” ） age(X, 35) ∧income（X , “4000-4900” ）→buys（X , “color TV” ） 合并为: age(X, “34-35”) ∧income（X , “3000-4900” ）→buys（X , “color TV” ） ;Mining Quantitative Associations;Mining Quantitative Associations;Correlation Analysis;Correlation Analysis;Correlation and Lift;Correlation Analysis举例;Correlation rule Mining;Constraint-based Association Mining;Metarule-Guided Mining of AR;Metarule-Guided Mining of AR;Mining Guided by Additional Rule Constraint;在其上的一个查询: Mine associations as Live(C,_, “Vancouver”) ∧Sales(C,?{I},{S}) → Sales(C,?{J},{T}) From Sales Where S.year=1999 ∧T.year=1999 ∧I.category=J.category Group by C, I.category Having sum(I.price) ≤ 100 ∧min(J.price) ≥500 With support threshold=0.01 With confidence threshold=0.5 ;Rule constraints classification;Query Optimization for Constrained FPMA;Anti-Monotonicity;Anti-Monotonicity;Anti-monotonic Constraints;Monotonicity;Monotonic Constraints;Succinctness;Succinct Constraints;Converting “Tough” Constraints;Strongly Convertible Constraints;A General Picture of Constraints;Classification of Constraints