数据挖掘第二次作业.docVIP

数据挖掘第二次作业 1. Find all frequent itemsets using Apriori and FP-growth, respectively, by treating each transaction ID as a market basket. Compare the efficiency of the two mining processes. Apriori：最小支持度计数值10*50%=5 a 7 b 6 c 5 d 6 e 8 ab 4 ad 4 ac 3 ae 6 bd 2 bc 2 be 5 cd 2 de 5 ae 6 be 5 de 5 a 7 b 6 d 6 c 5 e 8 C1= L1= C2= L2= abe 3 ade 4 C3= 没有频繁3项级 L={{a}{b}{c}{d}{e}{ae}{be}{de}} FP-growth 数据库的第一次扫描同Apriori，导出频繁1项集的集合和支持度计数（最小支持度计数为5）。频繁1项集按支持度递减序排序，结果集记为：L={{e:8},{a:7},{b:6}, {d:6},{c:5}}。 8 a 7 b 6 d 6 c 5 构造FP树如下：Use the results in part (a) to compute the con?dence for the association rules {a, d}({e} and {e}({a, d}. Is con?dence a symmetric measure? {a, d}= {e}, confidence=4/4=100% {e}= {a, d}, confidence=4/8=50% 两个关联规则的置信度不相等，因此置信度不是对称规则。 List all of the strong association rules (with support s and confidence c) matching the following metarule, where X is a variable representing customers, and itemi denotes variables representing items (e.g. “A”, “B”, etc.): 由(a)中得出的频繁项集L可得所有的关联规则如下： 对{a, e}有两个子集{a},{e}, 得到的关联规则为： a=e, confidence=6/7=86% e=a, confidence=6/8=75% 对{b, e}有两个子集{b}, {e},得到的关联规则为： b=e, confidence=5/6=83% e=b, confidence=5/8=62.5% 对{d, e}有两个子集{d}, {e},得到的关联规则为： d=e, confidence=5/6=83% e=d, confidence=5/8=62.5% 因为min_conf=60%，得出所有强关联规则为： buys(X, a)=buys(X, e) [s=60%, c=86%] buys(X, e)=buys(X, a) [s=60%, c=75%] buys(X, b)=buys(X, e) [s=50%, c=83%] buys(X, e)=buys(X, b) [s=50%, c=62.5%] buys(X, d)=buys(X, e) [s=50%, c=83%] buys(X, e)=buys(X, d) [s=50%, c=62.5%] 2. Compute the Information Gain for Gender, Car Type and Shirt Size. 从属性Gender开始 其余同理可得： Construct a decision tree with Information Gain. Family luxury? sports? Small? Not small? Not large? Large? YES? NO? Design a multilayer feed-forward neural network (one hidden layer) for the given data. Label th