Data Mining Algorithm

Large amounts of data are being generated and stored every day in Organizational computer database systems. Data mining is to discover knowledge from large amounts of data and is widely used in business world. Mining association rules from transactional data is becoming a popular and important knowledge discovery technique. Association rule mining is a data mining task that discovers relationships among items in a transactional database. One of the branches of data mining is Associative Classification (AC). AC algorithms integrate association rules discovery and classification to build a classifier from a training data for predicting the class of unforeseen test data. AC algorithms typically build a classifier by discovering the full set of

The previously unknown knowledge mined increases business intelligence, provides better support for decision making and consequently promotes the business competition. In order to discover rich and useful knowledge, many different types of data mining techniques are used. Mining association rules [1] from transactional data is becoming a popular and important knowledge discovery technique. Association rule mining is a data mining task that discovers relationships among items in a transactional database. An association rule is an implication of the form A B, where A and B are frequent itemsets in a transaction database and A∩B = . In practical applications, the rule A B can be used to predict that ‘if A occurs in a transaction, then B will likely also occur in the same transaction’, and we can apply this association rule to place ‘B close to A’ in the store layout and product placement of supermarket management. Association rules have been extensively studied in the literature for their usefulness in many application domains such as recommender systems, diagnosis decisions support, telecommunication, intrusion detection, etc. The efficient discovery of such rules has been a major focus in the data mining research

Moreover, the rule discovery process in traditional AC algorithms is not well integrated with the classification process.

2. Problem Definition

The associative classifier is a classifier that uses association rule mining in the training phase in order to generate classification rules. To use this classifier, datasets have to be transformed in a transactional format. Considering each attribute-value pair in a dataset as an item results in a transactional dataset in which a row of data looks like a transaction of items. Among items of each transaction, one is the class label of the related object. Using an association rule mining technique on the resulting transactional data, frequent itemsets are mined and the ones of the form {A, c} are extracted where A is a set of features and c is a class label (A and c are disjoint subsets of items). Among these frequent itemsets, the confident ones are chosen to build classification rules of the form A c. Then, these rules are used to predict class labels for objects with an unknown class.

Given a training data set T, for a rule R :