You are trying to predict the total sales per customer based on demographic information (age, gender, etc.), psychographic information (value focused, health focused, etc.), and employment information (salary, job type, etc.). In total, your sample has 2 million customer observations and approximately 78 potential predictors of average sales. The problem is complex so you decide to try four different techniques to predict total sales.Below are the r square results.

TrainingValidationTestLinear0.40.390.38Gradient Boosting0.440.410.41Random Forest0.450.440.43Neural Network0.720.620.45

Which model would you choose as the best?

Select one:

a.Gradient Boosting

b.Linear

c.Neural Network

d.Random Forest