Machine Learning based Prediction of Customer Spending Score

Machine learning models are widely used for the prediction of customer response in diverse markets. A number of machine learning techniques and models exist, ranging from complex techniques based on deep learning to comparatively simpler techniques based on regression. The current paper deals with the use of linear regression to estimate customer spending score at a mall using metrics such as age and annual income. The standard linear estimation results are bettered using data grouping and spline-based piecewise linear estimation. Further improvement in the root mean square error is achieved using adaptive splines using initial error estimates for more accurate prediction of user spending scores.