NEW DISCOVERY OF ASSOCIATION RULES IN DATA MINING THAT IS IMPLEMENTED IN THE REAL LIFE

NEW DISCOVERY OF ASSOCIATION RULES IN DATA MINING THAT IS IMPLEMENTED IN THE REAL LIFE

What is Association Rules?

According to blog of Margaret Rouse, Association rules are if/then statements that help uncover relationships between seemingly unrelated data in a relational database or other information repository. An example of an association rule would be "If a customer buys a dozen eggs, he is 80% likely to also purchase milk."

An association rule has two parts, an antecedent (if) and a consequent (then). An antecedent is an item found in the data. A consequent is an item that is found in combination with the antecedent. Association rules are created by analyzing data for frequent if/then patterns and using the criteria support and confidence to identify the most important relationships. Support is an indication of how frequently the items appear in the database. Confidence indicates the number of times the if/then statements have been found to be true.

Association Rules to Predict the Weather

I would like to create a dummy data for BMKG that can be used to predict the weather and this kind of information can be shared to the citizens. Given a set of data or information, that will predict the occurrence of an item based on the occurrences of other items

TID	ITEMS
1	Storm, Rainy, Thunder, Drizzle
2	Thunder, Rainy, Windy, Storm, Drizzle
3	Windy, Rainy, Drizzle, Storm
4	Drizzle, Rainy, Thunder
5	Thunder
6	Thunder, Drizzle
7	Rainy, Thunder, Storm
8	Drizzle, Windy

Mining Association Rules

There are two approaches of mining association rules:

1.      Frequent Itemset Generation

In order to generate a frequent itemset list one must avoid using the brute force approach because it can be very expensive to search through the whole data set to find the support count of each itemset. Some of the strategies that used to fix this problem are:

Ø  Reduce the number of candidates (Apriori Principle): use pruning techniques such as the Apriori principle to eliminate some of the candidate itemsets without counting their support values

Ø  Reduce the number of transactions: by combining transactions together we can reduce the total number of transactions

Ø  Reduce the number of comparisons (FP-Growth): use efficient data structures to store the candidates thereby eliminating the need to match every candidate against every transaction.

2.      Rule Generation

Generate high confidence rules from each frequent itemset, where each rule is a binary partitioning of a frequent itemset. Frequent itemset generation is still computationally expensive

Predict the Dummy Data of the Weather by Using Mining Association Rules

In this case, I would like to use the first Mining Association Rules to predict and analyze the dummy data of the weather. In the first rule there are three strategies that can be fixed to slolve the problem and I will only use two strategies (Apriori Principle and FP-Growth).

The Dummy Data (I give an example here for minimum support is 30%)

TID	ITEMS
1	Storm, Rainy, Thunder, Drizzle
2	Thunder, Rainy, Windy, Storm, Drizzle
3	Windy, Rainy, Drizzle, Storm
4	Drizzle, Rainy, Thunder
5	Thunder
6	Thunder, Drizzle
7	Rainy, Thunder, Storm
8	Drizzle, Windy

I give an example here for minimum support is 30%

Calculate minimum support

30% * 8 = 2.4

Frequency of occurrence

Items	Frequency
A	5
B	6
C	3
D	6
E	4

Prioritize the Item

Items	Frequency	Priority
A	5	3
B	6	1
C	3	5
D	6	2
E	4	4

Order the items according to the priority

TID	Items	Ordered Items
1	Storm, Rainy, Thunder, Drizzle	Drizzle, Thunder, Rainy, Storm
2	Thunder, Rainy, Windy, Storm, Drizzle	Drizzle, Thunder, Rainy, Storm, Windy
3	Windy, Rainy, Drizzle, Storm	Drizzle, Rainy, Storm, Windy
4	Drizzle, Rainy, Thunder	Drizzle, Thunder, Rainy
5	Thunder	Thunder
6	Thunder, Drizzle	Drizzle, Thunder
7	Rainy, Thunder, Storm	Thunder, Rainy, Storm
8	Drizzle, Windy	Drizzle, Windy

FP-Tree

Validation

REFERENCES

http://searchbusinessanalytics.techtarget.com/definition/association-rules-in-data-mining

http://www.belajaringgris.net/weather-vocabulary-3252.html

http://www.hypertextbookshop.com/dataminingbook/working_version/contents/chapters/chapter002/section002/blue/page001.html