Survey on Classification Techniques for Mining Temporal Patterns

Depiction of temporal information and examination of temporal knowledge are essential for effective temporal reasoning in medical analysis systems which are attaining significance recently. These temporal representation models afford attributes for temporal rules mining through association rule extraction which are necessary to perform detective inference in medical expert systems. Existing temporal mining systems focus on mining clinical data assumes instant stamping of tuples as the time stamping method. In such systems, database is created using a sequence of observations taken at various time slots and the mining algorithm classifies the data depending on the time instants recorded in the database. However, the analysis techniques need to consider data performing to a time interval. Hence, it is required to propose new depiction and reasoning methods for the effective analysis of medical data. In this paper, new temporal mining algorithms have been stated for extracting temporal patterns from clinical datasets.