Project 1 ‘Part 3’-‘Association’
Sachin Samal (ECU ID 250008)
Oct 21, 2021
DATA MINING WITH R
Here in this part of project, I have decided to perform the association analysis on my previous dataset from Project 1 and Project 2. For this project, I have pulled the data from English Premier League Results.
WHAT IS ASSOCIATION?
It is the method to discover relationships between seemingly independent relational databases or other data repositories. It aims to observe frequently occurring patterns, correlations, or associations from datasets found in various kinds of databases such as relational databases, transactional databases, and other forms of repositories.
LETS START OUT ASSOCIATION MINING…
LOADING MY EXCEL DATA INTO R ENVIRONMENT
library(readxl)
Results <- read_excel("Results.xlsx")
str(Results)## tibble [380 × 6] (S3: tbl_df/tbl/data.frame)
## $ Home_team: chr [1:380] "Arsenal" "Watford" "Chelsea" "Crystal Palace" ...
## $ Away_team: chr [1:380] "Leicester City" "Liverpool" "Burnley" "Huddersfield Town" ...
## $ Home_goal: num [1:380] 4 3 2 0 1 0 1 0 0 4 ...
## $ Away_goal: num [1:380] 3 3 3 3 0 0 0 2 2 0 ...
## $ Result : chr [1:380] "H" "D" "A" "A" ...
## $ Season : chr [1:380] "2017-2018" "2017-2018" "2017-2018" "2017-2018" ...summary(Results$Home_goal)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.000 1.000 1.000 1.532 2.000 7.000ASSOCIATION MINING WITH R
library(arules)## Loading required package: Matrix##
## Attaching package: 'arules'## The following objects are masked from 'package:base':
##
## abbreviate, writelibrary(arulesViz)Here, I have added some new libraries. Now, I need to convert the data for association analysis.
Now, I need to look at the columns to see if I can convert them into factors (or Boolean values) for analysis. This is because if not, I have to use different methods like “Kruskal Wallis test” or “Chi-square test” based on the requirement of my dataset.
colnames(Results)[c(1,2,3,4,5)]## [1] "Home_team" "Away_team" "Home_goal" "Away_goal" "Result"I don’t find anything that needs much attention. So , I will continue with the process as it needs.
CREATE TRANSACTION
I will let R do the default discretization to the rest of the data. This is because I could not come up with better cutoffs for what is left in the dataset.
library(ggplot2)
trans <- transactions(Results)## Warning: Column(s) 1, 2, 3, 4, 5, 6 not logical or factor. Applying default
## discretization (see '? discretizeDF').## Warning in discretize(x = c(3, 3, 3, 3, 0, 0, 0, 2, 2, 0, 4, 2, 1, 0, 0, : The calculated breaks are: 0, 0, 1, 6
## Only unique breaks are used reducing the number of intervals. Look at ? discretize for details.The conversion gives a warning because only discrete features (factor and logical) can be directly translated into items. Continuous features need to be discretized first.
summary(Results[5])## Result
## Length:380
## Class :character
## Mode :characterggplot(Results, aes(Home_goal)) + geom_histogram(fill='blue', color='darkred')## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
table(Results$Home_goal)##
## 0 1 2 3 4 5 7
## 90 126 91 35 23 14 1library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:arules':
##
## intersect, recode, setdiff, setequal, union## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionResults <- Results %>% mutate(Home_goal = Home_goal > 0)
ggplot(Results, aes(Home_goal)) + geom_bar(fill='darkred', color='blue')
table(Results$Home_goal)##
## FALSE TRUE
## 90 290The condition we had was if (Home_goal > 0). From the table, we can conclude that home goals are more frequently happening than away goals.
INSPECT TRANSACTION
Now, Lets run the transaction and see how the data has cleaned.
summary(trans)## transactions as itemMatrix in sparse format with
## 380 rows (elements/itemsets/transactions) and
## 49 columns (items) and a density of 0.122449
##
## most frequent items:
## Season=2017-2018 Away_goal=[1,6] Result=H Home_goal=[2,7]
## 380 244 173 164
## Away_goal=[0,1) (Other)
## 136 1183
##
## element (itemset/transaction) length distribution:
## sizes
## 6
## 380
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 6 6 6 6 6 6
##
## includes extended item information - examples:
## labels variables levels
## 1 Home_team=AFC Bournemouth Home_team AFC Bournemouth
## 2 Home_team=Arsenal Home_team Arsenal
## 3 Home_team=Brighton and Hove Albion Home_team Brighton and Hove Albion
##
## includes extended transaction information - examples:
## transactionID
## 1 1
## 2 2
## 3 3library(colorRamps)
#plotting image for transaction
image(trans, fill="red")
THE MOST FREQUENT ITEMS ON THE DATASET
frequentItems <- apriori(trans, parameter=list(target = "frequent"))## Apriori
##
## Parameter specification:
## confidence minval smax arem aval originalSupport maxtime support minlen
## NA 0.1 1 none FALSE TRUE 5 0.1 1
## maxlen target ext
## 10 frequent itemsets TRUE
##
## Algorithmic control:
## filter tree heap memopt load sort verbose
## 0.1 TRUE TRUE FALSE TRUE 2 TRUE
##
## Absolute minimum support count: 38
##
## set item appearances ...[0 item(s)] done [0.00s].
## set transactions ...[49 item(s), 380 transaction(s)] done [0.00s].
## sorting and recoding items ... [9 item(s)] done [0.00s].
## creating transaction tree ... done [0.00s].
## checking subsets of size 1 2 3 4 done [0.00s].
## sorting transactions ... done [0.00s].
## writing ... [53 set(s)] done [0.00s].
## creating S4 object ... done [0.00s].inspect(frequentItems)## items support transIdenticalToItemsets count
## [1] {Home_goal=[0,1)} 0.2368421 0.0000000 90
## [2] {Result=D} 0.2605263 0.0000000 99
## [3] {Result=A} 0.2842105 0.0000000 108
## [4] {Home_goal=[1,2)} 0.3315789 0.0000000 126
## [5] {Away_goal=[0,1)} 0.3578947 0.0000000 136
## [6] {Home_goal=[2,7]} 0.4315789 0.0000000 164
## [7] {Result=H} 0.4552632 0.0000000 173
## [8] {Away_goal=[1,6]} 0.6421053 0.0000000 244
## [9] {Season=2017-2018} 1.0000000 0.0000000 380
## [10] {Home_goal=[0,1),
## Result=A} 0.1526316 0.0000000 58
## [11] {Home_goal=[0,1),
## Away_goal=[1,6]} 0.1526316 0.0000000 58
## [12] {Home_goal=[0,1),
## Season=2017-2018} 0.2368421 0.0000000 90
## [13] {Home_goal=[1,2),
## Result=D} 0.1184211 0.0000000 45
## [14] {Away_goal=[1,6],
## Result=D} 0.1763158 0.0000000 67
## [15] {Result=D,
## Season=2017-2018} 0.2605263 0.0000000 99
## [16] {Away_goal=[1,6],
## Result=A} 0.2842105 0.0000000 108
## [17] {Result=A,
## Season=2017-2018} 0.2842105 0.0000000 108
## [18] {Home_goal=[1,2),
## Away_goal=[0,1)} 0.1157895 0.0000000 44
## [19] {Home_goal=[1,2),
## Result=H} 0.1157895 0.0000000 44
## [20] {Home_goal=[1,2),
## Away_goal=[1,6]} 0.2157895 0.0000000 82
## [21] {Home_goal=[1,2),
## Season=2017-2018} 0.3315789 0.0000000 126
## [22] {Home_goal=[2,7],
## Away_goal=[0,1)} 0.1578947 0.0000000 60
## [23] {Away_goal=[0,1),
## Result=H} 0.2736842 0.0000000 104
## [24] {Away_goal=[0,1),
## Season=2017-2018} 0.3578947 0.0000000 136
## [25] {Home_goal=[2,7],
## Result=H} 0.3394737 0.0000000 129
## [26] {Home_goal=[2,7],
## Away_goal=[1,6]} 0.2736842 0.0000000 104
## [27] {Home_goal=[2,7],
## Season=2017-2018} 0.4315789 0.0000000 164
## [28] {Away_goal=[1,6],
## Result=H} 0.1815789 0.0000000 69
## [29] {Result=H,
## Season=2017-2018} 0.4552632 0.0000000 173
## [30] {Away_goal=[1,6],
## Season=2017-2018} 0.6421053 0.0000000 244
## [31] {Home_goal=[0,1),
## Away_goal=[1,6],
## Result=A} 0.1526316 0.0000000 58
## [32] {Home_goal=[0,1),
## Result=A,
## Season=2017-2018} 0.1526316 0.0000000 58
## [33] {Home_goal=[0,1),
## Away_goal=[1,6],
## Season=2017-2018} 0.1526316 0.0000000 58
## [34] {Home_goal=[1,2),
## Away_goal=[1,6],
## Result=D} 0.1184211 0.0000000 45
## [35] {Home_goal=[1,2),
## Result=D,
## Season=2017-2018} 0.1184211 0.0000000 45
## [36] {Away_goal=[1,6],
## Result=D,
## Season=2017-2018} 0.1763158 0.0000000 67
## [37] {Away_goal=[1,6],
## Result=A,
## Season=2017-2018} 0.2842105 0.0000000 108
## [38] {Home_goal=[1,2),
## Away_goal=[0,1),
## Result=H} 0.1157895 0.0000000 44
## [39] {Home_goal=[1,2),
## Away_goal=[0,1),
## Season=2017-2018} 0.1157895 0.0000000 44
## [40] {Home_goal=[1,2),
## Result=H,
## Season=2017-2018} 0.1157895 0.0000000 44
## [41] {Home_goal=[1,2),
## Away_goal=[1,6],
## Season=2017-2018} 0.2157895 0.0000000 82
## [42] {Home_goal=[2,7],
## Away_goal=[0,1),
## Result=H} 0.1578947 0.0000000 60
## [43] {Home_goal=[2,7],
## Away_goal=[0,1),
## Season=2017-2018} 0.1578947 0.0000000 60
## [44] {Away_goal=[0,1),
## Result=H,
## Season=2017-2018} 0.2736842 0.0000000 104
## [45] {Home_goal=[2,7],
## Away_goal=[1,6],
## Result=H} 0.1815789 0.0000000 69
## [46] {Home_goal=[2,7],
## Result=H,
## Season=2017-2018} 0.3394737 0.0000000 129
## [47] {Home_goal=[2,7],
## Away_goal=[1,6],
## Season=2017-2018} 0.2736842 0.0000000 104
## [48] {Away_goal=[1,6],
## Result=H,
## Season=2017-2018} 0.1815789 0.0000000 69
## [49] {Home_goal=[0,1),
## Away_goal=[1,6],
## Result=A,
## Season=2017-2018} 0.1526316 0.1526316 58
## [50] {Home_goal=[1,2),
## Away_goal=[1,6],
## Result=D,
## Season=2017-2018} 0.1184211 0.1184211 45
## [51] {Home_goal=[1,2),
## Away_goal=[0,1),
## Result=H,
## Season=2017-2018} 0.1157895 0.1157895 44
## [52] {Home_goal=[2,7],
## Away_goal=[0,1),
## Result=H,
## Season=2017-2018} 0.1578947 0.1578947 60
## [53] {Home_goal=[2,7],
## Away_goal=[1,6],
## Result=H,
## Season=2017-2018} 0.1815789 0.1815789 69#calculating the frequent items
frequentItems## set of 53 itemsetsggplot(tibble(`Itemset Size` = factor(size(frequentItems))), aes(`Itemset Size`)) + geom_bar(fill = "purple", color = "black")
Apriori WITH R - GENERATING RULES
We will generate parameters support and confidence for rule mining and lift for interestingness evaluation.
Support indicates how frequently the itemset appears in the dataset.
Confidence is the proportion of the true positive of the rule.
Lets find out the rules using the apriori algorithm.
library(arules)
#association rules.
rules <- apriori(Results,
parameter = list(supp = 0.05, conf = 0.9,
target = "rules"))## Warning: Column(s) 1, 2, 4, 5, 6 not logical or factor. Applying default
## discretization (see '? discretizeDF').## Warning in discretize(x = c(3, 3, 3, 3, 0, 0, 0, 2, 2, 0, 4, 2, 1, 0, 0, : The calculated breaks are: 0, 0, 1, 6
## Only unique breaks are used reducing the number of intervals. Look at ? discretize for details.## Apriori
##
## Parameter specification:
## confidence minval smax arem aval originalSupport maxtime support minlen
## 0.9 0.1 1 none FALSE TRUE 5 0.05 1
## maxlen target ext
## 10 rules TRUE
##
## Algorithmic control:
## filter tree heap memopt load sort verbose
## 0.1 TRUE TRUE FALSE TRUE 2 TRUE
##
## Absolute minimum support count: 19
##
## set item appearances ...[0 item(s)] done [0.00s].
## set transactions ...[47 item(s), 380 transaction(s)] done [0.00s].
## sorting and recoding items ... [47 item(s)] done [0.00s].
## creating transaction tree ... done [0.00s].
## checking subsets of size 1 2 3 4 done [0.00s].
## writing ... [77 rule(s)] done [0.00s].
## creating S4 object ... done [0.00s].The Apriori algorithm generated 10 rules with the given constraints (parameters). Lets dive into the Parameter Specification section of the output.
minval is the minimum value of the support an itemset should satisfy to be a part of a rule.
smax is the maximum support value for an itemset.
arem is an Additional Rule Evaluation Parameter (similar to lift).
aval is a logical indicating whether to return the additional rule evaluation measure selected with arem.
originalSupport is the traditional support value that consider both LHS and RHS items for calculating support. If you want to use only the LHS items for the calculation then you need to set this to FALSE.
maxtime is the maximum amount of time allowed to check for subsets.
minlen is the minimum number of items required in the rule.
maxlen is the maximum number of items that can be present in the rule.
#length
length(rules)## [1] 77#sorting the rules and printing it
rules.sorted <-sort(rules, by="lift")
inspect(rules.sorted)## lhs rhs support confidence coverage lift count
## [1] {Home_goal,
## Away_goal=[0,1)} => {Result=H} 0.27368421 1 0.27368421 2.196532 104
## [2] {Home_goal,
## Away_goal=[0,1),
## Season=2017-2018} => {Result=H} 0.27368421 1 0.27368421 2.196532 104
## [3] {Result=A} => {Away_goal=[1,6]} 0.28421053 1 0.28421053 1.557377 108
## [4] {Home_goal,
## Result=D} => {Away_goal=[1,6]} 0.17631579 1 0.17631579 1.557377 67
## [5] {Home_goal,
## Result=A} => {Away_goal=[1,6]} 0.13157895 1 0.13157895 1.557377 50
## [6] {Result=A,
## Season=2017-2018} => {Away_goal=[1,6]} 0.28421053 1 0.28421053 1.557377 108
## [7] {Home_goal,
## Result=D,
## Season=2017-2018} => {Away_goal=[1,6]} 0.17631579 1 0.17631579 1.557377 67
## [8] {Home_goal,
## Result=A,
## Season=2017-2018} => {Away_goal=[1,6]} 0.13157895 1 0.13157895 1.557377 50
## [9] {Result=H} => {Home_goal} 0.45526316 1 0.45526316 1.310345 173
## [10] {Away_goal=[1,6],
## Result=D} => {Home_goal} 0.17631579 1 0.17631579 1.310345 67
## [11] {Away_goal=[0,1),
## Result=H} => {Home_goal} 0.27368421 1 0.27368421 1.310345 104
## [12] {Away_goal=[1,6],
## Result=H} => {Home_goal} 0.18157895 1 0.18157895 1.310345 69
## [13] {Result=H,
## Season=2017-2018} => {Home_goal} 0.45526316 1 0.45526316 1.310345 173
## [14] {Away_goal=[1,6],
## Result=D,
## Season=2017-2018} => {Home_goal} 0.17631579 1 0.17631579 1.310345 67
## [15] {Away_goal=[0,1),
## Result=H,
## Season=2017-2018} => {Home_goal} 0.27368421 1 0.27368421 1.310345 104
## [16] {Away_goal=[1,6],
## Result=H,
## Season=2017-2018} => {Home_goal} 0.18157895 1 0.18157895 1.310345 69
## [17] {} => {Season=2017-2018} 1.00000000 1 1.00000000 1.000000 380
## [18] {Away_team=Manchester City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [19] {Home_team=Huddersfield Town} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [20] {Home_team=Swansea City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [21] {Home_team=Newcastle United} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [22] {Away_team=Manchester United} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [23] {Home_team=Southampton} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [24] {Away_team=Chelsea} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [25] {Home_team=Stoke City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [26] {Away_team=Tottenham Hotspur} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [27] {Home_team=Burnley} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [28] {Home_team=Brighton and Hove Albion} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [29] {Home_team=AFC Bournemouth} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [30] {Home_team=Everton} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [31] {Home_team=Crystal Palace} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [32] {Away_team=Burnley} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [33] {Away_team=Brighton and Hove Albion} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [34] {Away_team=Liverpool} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [35] {Away_team=Crystal Palace} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [36] {Away_team=Southampton} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [37] {Away_team=West Bromwich Albion} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [38] {Away_team=Watford} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [39] {Home_team=Leicester City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [40] {Away_team=Arsenal} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [41] {Home_team=West Ham United} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [42] {Home_team=West Bromwich Albion} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [43] {Away_team=Swansea City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [44] {Home_team=Watford} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [45] {Home_team=Chelsea} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [46] {Away_team=AFC Bournemouth} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [47] {Away_team=Leicester City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [48] {Away_team=Everton} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [49] {Home_team=Liverpool} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [50] {Away_team=Huddersfield Town} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [51] {Away_team=Stoke City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [52] {Away_team=West Ham United} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [53] {Away_team=Newcastle United} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [54] {Home_team=Manchester United} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [55] {Home_team=Tottenham Hotspur} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [56] {Home_team=Manchester City} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [57] {Home_team=Arsenal} => {Season=2017-2018} 0.05000000 1 0.05000000 1.000000 19
## [58] {Result=D} => {Season=2017-2018} 0.26052632 1 0.26052632 1.000000 99
## [59] {Result=A} => {Season=2017-2018} 0.28421053 1 0.28421053 1.000000 108
## [60] {Away_goal=[0,1)} => {Season=2017-2018} 0.35789474 1 0.35789474 1.000000 136
## [61] {Result=H} => {Season=2017-2018} 0.45526316 1 0.45526316 1.000000 173
## [62] {Away_goal=[1,6]} => {Season=2017-2018} 0.64210526 1 0.64210526 1.000000 244
## [63] {Home_goal} => {Season=2017-2018} 0.76315789 1 0.76315789 1.000000 290
## [64] {Away_goal=[0,1),
## Result=D} => {Season=2017-2018} 0.08421053 1 0.08421053 1.000000 32
## [65] {Away_goal=[1,6],
## Result=D} => {Season=2017-2018} 0.17631579 1 0.17631579 1.000000 67
## [66] {Home_goal,
## Result=D} => {Season=2017-2018} 0.17631579 1 0.17631579 1.000000 67
## [67] {Away_goal=[1,6],
## Result=A} => {Season=2017-2018} 0.28421053 1 0.28421053 1.000000 108
## [68] {Home_goal,
## Result=A} => {Season=2017-2018} 0.13157895 1 0.13157895 1.000000 50
## [69] {Away_goal=[0,1),
## Result=H} => {Season=2017-2018} 0.27368421 1 0.27368421 1.000000 104
## [70] {Home_goal,
## Away_goal=[0,1)} => {Season=2017-2018} 0.27368421 1 0.27368421 1.000000 104
## [71] {Away_goal=[1,6],
## Result=H} => {Season=2017-2018} 0.18157895 1 0.18157895 1.000000 69
## [72] {Home_goal,
## Result=H} => {Season=2017-2018} 0.45526316 1 0.45526316 1.000000 173
## [73] {Home_goal,
## Away_goal=[1,6]} => {Season=2017-2018} 0.48947368 1 0.48947368 1.000000 186
## [74] {Home_goal,
## Away_goal=[1,6],
## Result=D} => {Season=2017-2018} 0.17631579 1 0.17631579 1.000000 67
## [75] {Home_goal,
## Away_goal=[1,6],
## Result=A} => {Season=2017-2018} 0.13157895 1 0.13157895 1.000000 50
## [76] {Home_goal,
## Away_goal=[0,1),
## Result=H} => {Season=2017-2018} 0.27368421 1 0.27368421 1.000000 104
## [77] {Home_goal,
## Away_goal=[1,6],
## Result=H} => {Season=2017-2018} 0.18157895 1 0.18157895 1.000000 69summary(rules)## set of 77 rules
##
## rule length distribution (lhs + rhs):sizes
## 1 2 3 4
## 1 48 18 10
##
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1.000 2.000 2.000 2.481 3.000 4.000
##
## summary of quality measures:
## support confidence coverage lift
## Min. :0.0500 Min. :1 Min. :0.0500 Min. :1.000
## 1st Qu.:0.0500 1st Qu.:1 1st Qu.:0.0500 1st Qu.:1.000
## Median :0.0500 Median :1 Median :0.0500 Median :1.000
## Mean :0.1683 Mean :1 Mean :0.1683 Mean :1.107
## 3rd Qu.:0.2737 3rd Qu.:1 3rd Qu.:0.2737 3rd Qu.:1.000
## Max. :1.0000 Max. :1 Max. :1.0000 Max. :2.197
## count
## Min. : 19.00
## 1st Qu.: 19.00
## Median : 19.00
## Mean : 63.95
## 3rd Qu.:104.00
## Max. :380.00
##
## mining info:
## data ntransactions support confidence
## Results 380 0.05 0.9#plot rules.sorted
plot(rules.sorted)## To reduce overplotting, jitter is added! Use jitter = 0 to prevent jitter.
plot(rules, method = "graph", measure = "lift", shading = "confidence", engine = "htmlwidget")Sort by support
rules <- apriori(trans, parameter = list(support = 0.05, confidence = 0.9))## Apriori
##
## Parameter specification:
## confidence minval smax arem aval originalSupport maxtime support minlen
## 0.9 0.1 1 none FALSE TRUE 5 0.05 1
## maxlen target ext
## 10 rules TRUE
##
## Algorithmic control:
## filter tree heap memopt load sort verbose
## 0.1 TRUE TRUE FALSE TRUE 2 TRUE
##
## Absolute minimum support count: 19
##
## set item appearances ...[0 item(s)] done [0.00s].
## set transactions ...[49 item(s), 380 transaction(s)] done [0.00s].
## sorting and recoding items ... [49 item(s)] done [0.00s].
## creating transaction tree ... done [0.00s].
## checking subsets of size 1 2 3 4 done [0.00s].
## writing ... [101 rule(s)] done [0.00s].
## creating S4 object ... done [0.00s].Lets inspect the top three rules sorted by confidence.
inspect(head(sort(rules, by = "confidence"), 3))## lhs rhs support confidence
## [1] {} => {Season=2017-2018} 1.00 1
## [2] {Home_team=Arsenal} => {Season=2017-2018} 0.05 1
## [3] {Away_team=Tottenham Hotspur} => {Season=2017-2018} 0.05 1
## coverage lift count
## [1] 1.00 1 380
## [2] 0.05 1 19
## [3] 0.05 1 19VISUALIZING ASSOCIATION RULES
plot(rules,jitter = 1)
plot(rules, shading = "order", color=c("darkred", "purple"))## To reduce overplotting, jitter is added! Use jitter = 0 to prevent jitter.
#Graph plot for items
plot(rules, method="graph", max=20, control=list(verbose = FALSE), colors=c("red", "green"))## Warning: Too many rules supplied. Only plotting the best 20 rules using lift
## (change control parameter max if needed)
plot(rules, method = "matrix", measure=c("support","confidence"))## Itemsets in Antecedent (LHS)
## [1] "{Away_goal=[0,1),Result=D,Season=2017-2018}"
## [2] "{Home_goal=[0,1),Away_goal=[0,1),Season=2017-2018}"
## [3] "{Home_goal=[0,1),Away_goal=[1,6],Season=2017-2018}"
## [4] "{Home_goal=[0,1),Result=D,Season=2017-2018}"
## [5] "{Home_goal=[1,2),Result=H,Season=2017-2018}"
## [6] "{Away_goal=[0,1),Result=D}"
## [7] "{Home_goal=[0,1),Away_goal=[0,1)}"
## [8] "{Away_goal=[1,6],Result=H,Season=2017-2018}"
## [9] "{Home_goal=[0,1),Away_goal=[1,6]}"
## [10] "{Home_goal=[1,2),Away_goal=[0,1),Season=2017-2018}"
## [11] "{Home_goal=[2,7],Away_goal=[0,1),Season=2017-2018}"
## [12] "{Home_goal=[0,1),Result=D}"
## [13] "{Home_goal=[1,2),Result=H}"
## [14] "{Away_goal=[1,6],Result=H}"
## [15] "{Home_goal=[1,2),Away_goal=[0,1)}"
## [16] "{Home_goal=[2,7],Away_goal=[0,1)}"
## [17] "{Result=A,Season=2017-2018}"
## [18] "{Home_goal=[0,1),Result=A,Season=2017-2018}"
## [19] "{Home_goal=[1,2),Result=D,Season=2017-2018}"
## [20] "{Home_goal=[2,7],Result=D,Season=2017-2018}"
## [21] "{Home_goal=[1,2),Result=A,Season=2017-2018}"
## [22] "{Result=A}"
## [23] "{Home_goal=[0,1),Result=A}"
## [24] "{Home_goal=[1,2),Result=D}"
## [25] "{Home_goal=[2,7],Result=D}"
## [26] "{Home_goal=[1,2),Result=A}"
## [27] "{}"
## [28] "{Home_team=Arsenal}"
## [29] "{Away_team=Tottenham Hotspur}"
## [30] "{Home_team=Manchester United}"
## [31] "{Away_team=West Ham United}"
## [32] "{Home_team=Swansea City}"
## [33] "{Away_team=Manchester United}"
## [34] "{Home_team=AFC Bournemouth}"
## [35] "{Away_team=Watford}"
## [36] "{Home_team=Burnley}"
## [37] "{Away_team=West Bromwich Albion}"
## [38] "{Home_team=Leicester City}"
## [39] "{Away_team=Brighton and Hove Albion}"
## [40] "{Home_team=Liverpool}"
## [41] "{Away_team=Crystal Palace}"
## [42] "{Home_team=Stoke City}"
## [43] "{Away_team=Arsenal}"
## [44] "{Home_team=Huddersfield Town}"
## [45] "{Away_team=Newcastle United}"
## [46] "{Home_team=Tottenham Hotspur}"
## [47] "{Away_team=Chelsea}"
## [48] "{Home_team=Manchester City}"
## [49] "{Away_team=Everton}"
## [50] "{Away_team=Southampton}"
## [51] "{Home_team=Newcastle United}"
## [52] "{Away_team=Manchester City}"
## [53] "{Home_team=Brighton and Hove Albion}"
## [54] "{Away_team=Leicester City}"
## [55] "{Home_team=Watford}"
## [56] "{Away_team=Liverpool}"
## [57] "{Home_team=Chelsea}"
## [58] "{Away_team=Burnley}"
## [59] "{Home_team=Crystal Palace}"
## [60] "{Away_team=Huddersfield Town}"
## [61] "{Home_team=Everton}"
## [62] "{Away_team=Stoke City}"
## [63] "{Home_team=Southampton}"
## [64] "{Away_team=Swansea City}"
## [65] "{Home_team=West Bromwich Albion}"
## [66] "{Away_team=AFC Bournemouth}"
## [67] "{Home_team=West Ham United}"
## [68] "{Home_goal=[0,1)}"
## [69] "{Result=D}"
## [70] "{Home_goal=[1,2)}"
## [71] "{Away_goal=[0,1)}"
## [72] "{Home_goal=[2,7]}"
## [73] "{Result=H}"
## [74] "{Away_goal=[1,6]}"
## [75] "{Away_goal=[1,6],Result=D}"
## [76] "{Away_goal=[1,6],Result=A}"
## [77] "{Home_goal=[1,2),Away_goal=[1,6]}"
## [78] "{Away_goal=[0,1),Result=H}"
## [79] "{Home_goal=[2,7],Result=H}"
## [80] "{Home_goal=[2,7],Away_goal=[1,6]}"
## [81] "{Home_goal=[0,1),Away_goal=[0,1),Result=D}"
## [82] "{Home_goal=[0,1),Away_goal=[1,6],Result=A}"
## [83] "{Home_goal=[1,2),Away_goal=[1,6],Result=D}"
## [84] "{Home_goal=[2,7],Away_goal=[1,6],Result=D}"
## [85] "{Home_goal=[1,2),Away_goal=[1,6],Result=A}"
## [86] "{Home_goal=[1,2),Away_goal=[0,1),Result=H}"
## [87] "{Home_goal=[2,7],Away_goal=[0,1),Result=H}"
## [88] "{Home_goal=[2,7],Away_goal=[1,6],Result=H}"
## Itemsets in Consequent (RHS)
## [1] "{Season=2017-2018}" "{Away_goal=[1,6]}" "{Result=H}"
## [4] "{Home_goal=[2,7]}" "{Away_goal=[0,1)}" "{Result=A}"
## [7] "{Result=D}" "{Home_goal=[0,1)}"
ASSOCIATION MINING ANALYSIS
Here, association rules can be used to understand the prediction about which sides either home or away, are going to be more successful in future based on their current perdormances and result. Understanding the association or co-occurence will help us plan what promo or recommendation we will need to give to organizer based on their current outcomes. Network analysis help further help us find more insight compared to if only we look at the rules individually.
CONCLUSION
This method can be modified and implemented in different ways, depending on the user’s interest. A deeper look into the outcomes can establish additional rules for a more detailed analysis. From this analysis, I have found that the proposed association rule for data mining can be effective to extract football tactics from the team’s individual performance.
More than 60% times, the home teams have won when they have scored goals >2, while the away teams have had draw or lost outcomes on the remaining part. Although the presented technique is not a sophisticated measure for establishing a general recommendation pattern in this dataset, it provides us with an underlying relationships between the teams and their goal differences. Such approach can also be incorporated in many activities, for instance in pitch analysis or a marketing campaign.
———————————————————————–END———————————————————————–