An approach to identify the strength of the decision tree algorithm by comparing various classifiers for customer relationship improvising

Patel Alex
Research Scholar, Rai University, Ahmedabad , IN

Akash Saxena
Research Scholar, Rai University, Ahmedabad , IN

Keywords: Decision Tree, Data mining, Comparison, Customers.

Abstract

Objectives: To Compare classification algorithm and to identify the Strength of Decision Tree algorithm. Methods/Statistical analysis: Current study is to compare various classification algorithms theoretically. There are various sources of such data that majorly help us to study the user’s pattern or access pattern. One of the important ways is Data Mining. Data Mining is an effective tool to derive useful data from unstructured and unmanaged data. To compare classification algorithms, it requires to compare estimated accuracy between the algorithms as we want the best algorithm for the dataset. By study and referring various papers various methods are compared. Findings: There are various methods comparing algorithms, but before we jump into that it requires to make the theoretical concept clear. We may not know best algorithm or best parameters to use for algorithms because we never knew which algorithm is going to give the best result for our dataset. Before one applies an algorithm to the dataset it requires to clear various terms, to help and decide which algorithm to use for specific problems. In this study, the result has been derived by comparison of various algorithms and its methods. Algorithms are compared on the basis of Data size, Accuracy, Speed of working, Type, Training time duration and its characteristics. Application/Improvements: Purpose of comparison is not to prove any model high or low, but to search the accurate model as per problem. This comparison is theoretical so for the improvement it requires to work with data to compare the results.

How to Cite

References

NA