数据科学与大数据分析项目练习-3将Apriori算法应用于R中提供的“Groceries”数据集

我们需要安装arules and arulesViz包。

项目要求：

生成频繁项目集满足下面条件：
– The minimum support threshold as 0.02
– The minimum length of the itemsets as 1
– The maximum length of the itemsets as 10

生成的关联规则满足下面条件：
– The minimum support threshold as 0.001
– The minimum confidence threshold as 0.6

Project Start

# 下载并加载相关的包（可能需要梯子）
install.packages('arules') 
install.packages('arulesViz')
library('arules')
library('arulesViz')
1
2
3
4
5

# 读取并展示“Groceries”包
data(Groceries)
Groceries
summary(Groceries)
class(Groceries)
1
2
3
4
5

可以看到这个“Groceries”包一共有9835行和169列。

# 展示前20个grocery labels
Groceries@itemInfo[1:20,]

# 显示第10至20的transactions
apply(Groceries@data[,10:20], 2, 
      function(r) paste(Groceries@itemInfo[r,"labels"], collapse=", ")
)
1
2
3
4
5
6
7

接下来是生成频繁项目集
参数的设置按照要求

# frequent 1-itemsets
itemsets <- apriori(Groceries, parameter=list(minlen=1, maxlen=1, support=0.02, target="frequent itemsets"))
summary(itemsets)
inspect(head(sort(itemsets, by = "support"), 10))

# frequent 2-itemsets
itemsets <- apriori(Groceries, parameter=list(minlen=2, maxlen=2, support=0.02, target="frequent itemsets"))
summary(itemsets)
inspect(head(sort(itemsets, by ="support"),10))

# frequent 3-itemsets
itemsets <- apriori(Groceries, parameter=list(minlen=3, maxlen=3, support=0.02, target="frequent itemsets"))
inspect(sort(itemsets, by ="support"))

# frequent 4-itemsets
itemsets <- apriori(Groceries, parameter=list(minlen=4, maxlen=4, support=0.02, target="frequent itemsets"))
inspect(sort(itemsets, by ="support"))
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

生成的频繁项目集结果如下：

规则生成和可视化

rules <- apriori(Groceries, parameter=list(support=0.001,
                                           confidence=0.6, target = "rules"))
summary(rules)
plot(rules)
plot(rules@quality)
1
2
3
4
5

# 显示rules与最高lift scores
inspect(head(sort(rules, by="lift"), 10))
1
2

筛选出置信度大于0.9的rules

confidentRules <- rules[quality(rules)$confidence > 0.9]
confidentRules

plot(confidentRules, method="matrix", measure=c("lift", "confidence"), control=list(recorder=TRUE))
1
2
3
4

一个127个，之后polt成为matrix得到

# 选择lift最高的5项规则
highLiftRules <- head(sort(rules, by="lift"), 5)
plot(highLiftRules, method="graph", control=list(type="items"))
1
2
3

得到图片如下所示。
从图中我们可以得到买火腿的基本上都会购买加工芝士，买爆米花或者soda饮料的很大概率会购买咸味小吃。

参考书目

1. Data Science and Big Data Analytics: Discovering, Analyzing, Visualizing and Presenting Data, EMC Education Services, John Wiley & Sons, 27 Jan. 2015