chapter  14
Discrete Sequence Classification
ByMohammad Al Hasan
Pages 19

Sequence classification is an important research task with numerous applications in various domains, including bioinformatics, e-commerce, health informatics, computer security, and finance. In bioinformatics domain, classification of protein sequences is used to predict the structure and function of newly discovered proteins [12]. In e-commerce, query sequences from a session are used to distinguish committed shoppers from the window shoppers [42]. In the Web application domain, a sequence of queries is used to distinguish Web robots from human users [49]. In health informatics, sequence of events is used in longitudinal studies for predicting the risk factor of a patient for a given disease [17]. In the security domain, the sequence of a user’s system activities on a terminal is monitored for detecting abnormal activities [27]. In the financial world, sequences of customer activities, such as credit card transactions, are monitored to identify potential identity theft, sequences of bank transactions are used for classifying accounts that are used for money laundering [33], and so on. See [36] for a survey on sequential pattern mining