Volume : VI, Issue : XI, November - 2017

Improvement in Random Forest Regression Model using a Growth Predictor

Bhavna Saluja, Simmi Saluja

Abstract :

 Customer data analytics refers to the systematic study of the company’s customer information that helps make key business decisions. By market segmentation and predictive analysis, we can identify and retain the profitable customers, locate sites for a successful business and understand customer relationship management. Analysing user intentionality towards products is crucial for retargeting. Predictive models give a numerical measure which is linked with each client for their propensity to churn and the result is in terms of probability. This information proves to be very useful in developing marketing campaigns aimed at customer retention. The research in this work is based on the empirical results obtained by making regression models on consumer retail data sets. Random Forest algorithm is one of the best algorithms that is used today for this purpose. In this paper, an attempt has been made to improve the accuracy of the model through introduction of a metric– growth variable– associated with predictor variables. It is also discussed how coverage or accuracy of the Random Forest model can be improved by careful selection of the variables being used to build the model and looking at the data from different perspectives and summarizing the relationships identified. Experimental results clearly indicate the enhancement in Random Forest model in terms of accuracy.

Keywords :

Article: Download PDF   DOI : 10.36106/ijsr  

Cite This Article:

Bhavna Saluja, Simmi Saluja, Improvement in Random Forest Regression Model using a Growth Predictor, INTERNATIONAL JOURNAL OF SCIENTIFIC RESEARCH : Volume-6 | Issue-11 | November-2017

Number of Downloads : 436

References :