Examples Of Data Mining In Healthcare

In this fast paced competitive world, human health is considered to be priceless. Health once lost is difficult to be recovered. Therefore a sincere attempt is made to effectively incorporate the benefits of information technology for healthcare to make the wellbeing of humans a priority.

Healthcare industry consists of humungous amount of data. A methodical procedure for analyzing, storing, processing and validating this data is necessary. Therefore to achieve this goal, major techniques like data mining and hadoop have contributed various forms to deliver applications in the area of healthcare. WEKA is a collection of machine learning algorithms that can be used for data mining tasks in healthcare. However, analyzing healthcare data using

Domains can be bank transactions, scientific data , medical and personal data, surveillance video and picture, satellite sensing, world wide web repositories etc. It then continues with data selection, where the target data set of the desired domain is taken into consideration. The next step includes data preprocessing which is followed by application of analytical method on the preprocessed data. Analytical method can be either data mining or hadoop. The output pattern is then evaluated for knowledge discovery .The final output is displayed in an interface viewable by the common man. Figure illustrates the

It is an important step of KDD process.

Figure lists the following iterative sequential steps

1. Data collection: learning the application domain. Eg: Medical and personal data

2. Data selection: creating a target data set which will be subjected to analysis. Eg: heart disease dataset from the UCI repository

3. Data pre-processing: The chosen health care datasets are pre-processed to handle problems like noise, missing and inconsistent data. This step transforms data into a form that is presentable to the data mining techniques.

4. Data mining: This involves the task of analyzing the dataset and extracting the data patterns using various data mining algorithms like classification, regression, association and clustering.

5. Pattern evaluation and knowledge discovery: A systematic determination of strictly interesting patterns representing knowledge, is done using criteria governed by a set of standards.

6. Result visualization: It is a final Phase where the knowledge discovered is represented visually to the user to help understand and interpret the