Data mining is the approach which can extract useful information from the data. The prediction analysis is the approach which can predict future possibilities based on the current information. The authors propose a hybrid classifier to carry out the heart disease prediction. The hybrid classifier is combination of random forest and decision tree classifier. Moreover, the heart disease prediction technique has three steps, which are data pre-processing, feature extraction, and classification. In this research, random forest classifier is applied for the feature extraction and decision tree classifier is applied for the generation of prediction results. However, random forest classifier will extract the information and decision tree will generate final classifier result. The authors show the results of proposed model using the Python platform. Moreover, the results are compared with support vector machine (SVM) and k-nearest neighbour classifier (KNN).
TopData Mining
In archives, libraries and even other appliances, there is a vast volume of data being stored, since personal and private data cannot be stored somewhere else. Therefore, a method in which both data and information (Dey & Rautaray, 2014) can be safely and securely stored is very important to discover and suggest. Users often find it very difficult to collect and use only valuable knowledge from this tremendous data. Data mining is therefore used to solve this case. Data mining is the method of only sorting, selecting and processing data which is valuable and important for that unique moment of time. This helps the user to view their details from anywhere and at any time. (Oyelade, Oladipupo, & Obagbuwa, 2010).
The files and many other areas are stored with a vast volume of appropriate and irrelevant data. It has given attention to the word data mining, which in the decision-making process can be more useful. It is step of gathering meaningful and significant data from vast volume of data that is collected on the internet nearly everywhere.
It consists of steps in a repetitive sequence:
- 1.
Data cleaning: This reduces unnecessary noise and inconsistent information.
- 2.
Data Integration: By this process, different data sources are integrated.
- 3.
Data Selection: By this stage, entirely different data is retrieved from databases. (Rauf, Khusro, Javed, & Saeed, 2012).
- 4.
Data Processing: The transformation of data in which summary or grouping operations are conducted is carried out in a highly suitable fashion.
- 5.
Data mining: Data is collected in many ways, and is considered to be a very significant step.
- 6.
Information Presentation: Here, with the assistance of multiple representation and visualization methods, the data that is mined is resented to the user.
The logical process that was used to search for interesting and specific data from vast volumes of stored data is data mining. To find the patterns that were already used and established, this method is used. As these trends are identified, they are further used to make decisions for the development of markets and enterprises (Desai, 2013).
Figure 1. Data Mining as KDD Process
There are 3 phases involved:
- •
Exploration: In the initial stage, the cleaning and transfer of information is performed in a different fashion. This results in the resolution of the underlying questions associated with the form of data obtained.
- •
Identification of patterns: As the data is explored, optimized and resolved for a particular reason, a pattern for identification is created. In this, a trend that makes the best forecast is found and chosen.(Taneja, 2013).
- •
Deployment: To have a desired effect, various patterns are implemented in this phase.