KDD Techniques for Diagnosis of Cardiac Data Sets

Abstract

Now-a-days, data is growing at a phenomenal rate from terabytes to petabytes. This growth is reflected in an increase in both the size and complexity as individual data base as well as in a proliferation of new databases. Raw data stored in databases are seldom of direct use. In practical applications, data are usually presented to the users in a modified form, tailored to satisfy specific needs. Despite this, people must analyze data more or less manually, acting as sophisticated query processors. This may be satisfactory if the total amount of data being analyzed is relatively small, but is unacceptable for large amounts of data. The healthcare domain is generally perceived as being information rich , however, there is lack of effective analysis tools to discover hidden relationships and trends in data. An automation of data analysis task is required in such cases. Knowledge Discovery in Database (KDD) is an emerging field that aims at analyzing massive amounts of (extracts) meaningful and comprehensible patterns, called knowledge. Formal definition for Knowledge discovery in databases (KDD) is defined as the nontrivial process of identifying valid, novel, potentially useful and ultimately understandable patterns in data newlineThe phases of the KDD process are as follows. Selection is an appropriate procedure for generating the target data set from the database. Its major goal is to select typical data from the database, in order to make the target data set as representative as possible. In most practical cases, data in the database and in the target data set contain noise, i.e. erroneous, inexact, imprecise, conflicting, exceptional and missing values including ambiguities. Phases of pre-processing eliminate noise from the target data set and possibly generate specific data sequences in the set of pre-processed data. The next phase is the transformation of the pre-processed data into a suitable form for performing the desired Data Mining (DM) task. Typically, transformations include some appropriate reduction