Data Preprocessing Methods and Unified Framework for a Cardiac Database

Data in the real world is dirty. If the data in not in the quality then it is not possible to get the correct result in the mining. There are some measures for data qualities. Accuracy, Completeness, Consistency, Timeliness, Believability, Value added, Interpretability, and Accessibility are the measures. To have these measures preprocessing in very important. In this paper we have proposed the improved algorithms for data cleaning, integration, and discretization. Thus our goal is to provide best suited algorithms for data cleaning of the cardiology data set.