In order to quantify the certainty of our estimate, we will use bootstrapping method. We will resample the original data used to train the neural network by randomly select observations from the original dataset with replacement. Using these new datasets, we create new neural networks and train them in the same way as the original model. Then, we take all the new estimates and construct a new distribution by using the new model to predict the spend for consumers who are in their 40's. From this distribution, we can construct a confidence interval (say, 95%) that gives us a range of values that we can be reasonably certain will contain the true average spend for consumers in their 40s