JarCastingjava-frequent-pattern-mining
Package provides java implementation of frequent pattern mining algorithms such as apriori, fp-growth
Features
- Apriori
- FP-Growth
Install
Add the following dependency to your POM file:
<dependency>
<groupId>com.github.chen0040</groupId>
<artifactId>java-frequent-pattern-mining</artifactId>
<version>1.0.1</version>
</dependency>
Usage
FP Growth
The sample code below shows how to use FPGrowth to obtain the frequent item sets from a list of transactions
List<List<String>> database = new ArrayList<>();
// each line below add a new transaction to the database
database.add(Arrays.asList("f", "a", "c", "d", "g", "i", "m", "p"));
database.add(Arrays.asList("a", "b", "c", "f", "l", "m", "o"));
database.add(Arrays.asList("b", "f", "h", "j", "o", "w"));
database.add(Arrays.asList("b", "c", "k", "s", "p"));
database.add(Arrays.asList("a", "f", "c", "e", "l", "p", "m", "n"));
AssocRuleMiner method = new FPGrowth();
method.setMinSupportLevel(2);
MetaData metaData = new MetaData(database);
// obtain all frequent item sets with support level not below 2
ItemSets frequent_item_sets = method.minePatterns(database, metaData.getUniqueItems());
fis.stream().forEach(itemSet -> System.out.println("item-set: " + itemSet));
// obtain the max frequent item sets
ItemSets max_frequent_item_sets = method.findMaxPatterns(database, metaData.getUniqueItems());
Apriori
The sample code below shows how to use Apriori to obtain the frequent item sets from a list of transactions
List<List<String>> database = new ArrayList<>();
// each line below add a new transaction to the database
database.add(Arrays.asList("f", "a", "c", "d", "g", "i", "m", "p"));
database.add(Arrays.asList("a", "b", "c", "f", "l", "m", "o"));
database.add(Arrays.asList("b", "f", "h", "j", "o", "w"));
database.add(Arrays.asList("b", "c", "k", "s", "p"));
database.add(Arrays.asList("a", "f", "c", "e", "l", "p", "m", "n"));
AssocRuleMiner method = new Apriori();
method.setMinSupportLevel(2);
MetaData metaData = new MetaData(database);
// obtain all frequent item sets with support level not below 2
ItemSets frequent_item_sets = method.minePatterns(database, metaData.getUniqueItems());
fis.stream().forEach(itemSet -> System.out.println("item-set: " + itemSet));
// obtain the max frequent item sets
ItemSets max_frequent_item_sets = method.findMaxPatterns(database, metaData.getUniqueItems());