Data Mining Process

1.      Through data analysis, pattern identification, as well as relationship, developed to solve the problems this process is known as data mining sorting by large sets of data. To predict future trends, data mining tools allow enterprises. There is another definition of data mining is that previously unknown patterns in data that were discovered are known as data mining.

2.      To increase sales by using the data mining popularity, data mining has many definitions because beyond those limits it has been stretched to include most forms of data analysis through some software vendors.

3.      For the recent popularity of data mining there are many reasons here we describe some of them.

·         In large data source, untapped hidden value is recognized in general.

·         In the form of a data warehouse database consolidation as well as into a single location other data repositories.

·         In data processing as well as storage technologies the exponential increase.

4.      To purchase data mining software before making a decision organization should follow some standard criteria in any important software: analysis of cost, to use the software people with expertise perform the data analyses, historical data availability, for the data mining software a business need.

 5.      In large datasets, a hidden pattern discovers and identify by data mining. In Data analysis this is one of the activities. On structured data mostly studied in data mining. Data mining is a merged or multiple blend disciplines otherwise analytical tools as well as techniques.

6.      Data mining have three main methods such as clustering, association as well as prediction. Prediction is that type of data mining which tells us about the future. There are further two types of prediction Classification or regression. With similar characteristics finding the group entities is known as clustering. Association is that type of method in which items occur together, and the relationship is established.

7.      Following application are listed such as banking, healthcare, entertainment, computer hardware, and software as well as sports. For prediction as well as forecasting the commonalities are the need for planning purposes as well as also support in decision making.

8.      A general process is usually followed to carry out the mining project. To be successful in any scheme of data mining must monitor systematic project management. CRISP-DM, KDD as well as SEMMA are the numerous processes.

CRISP-DM takes inclusive techniques which include business understanding as well as the relevant data whereas SEMMA assumes the goals of the data mining project, as well as objective with the proper data sources, have been understood as well as identified.

9. In nature, these steps are sequential. There is usually a deal of backtracking. By experience and experimentation, data mining is driven which is dependent on the problem as well as analyst experience, time-consuming as well as the process could be iterative.

10. To any successful data mining study data preprocessing is significant. Because of the functional data, useful information is found, and the right decision is made. Four main steps data preprocessing includes: consolidation of data, cleaning of data, the transformation of data, as well as reduction of data.

         11. The assessment, as well as comparative analysis of several models, built the building model step also encompasses. Because for data mining task there is not a universally called best algorithm.

12. From past data, classification learns pattern into their respective classes or groups. The primary purpose of the rating has stored the database as well as model generate which predict the behavior of the future.  For solving the Classification problem cluster analysis is an exploratory data analysis tool. According to demographics, customers can be grouped.

13. In this chapter, we are talking about the data mining of the project. We discussed in this chapter about the methods of data mining. Cluster, classification along with prediction is the three methods of data mining that is used.

14. In data mining preservation of privacy has emerged for exchanging confidential an absolute prerequisite data in terms of validation, data analysis, as well as publishing. In data mining techniques the current privacy preserving is classified depending on the hide association rule, distortion as well as distributed.

15.  Crystal-ball prediction provides instant data mining.

·         For business applications, data mining is not yet viable

·         A separate, dedicated database data mining require these.