Question

Give two examples of business decisions that could benefit from each of AI's problem types (regression, categorization, and clustering)?

Answer 1

Regression: Suppose we wish to predict the significant drivers of sales. After determining variables such as price, assortment etc, we wish to determine what the future sales will be.

Suppose we wish to find out the significant drivers of a country's growth. A regression model can be run to check for significance of each of the economic variables under consideration.

Categorization:

Suppose we wish to see whether a person is a low risk or a high risk customer before issuing a credit card to him. Categorization will help in this case.

Suppose we wish to predict churn of a customer. This model can be used for effective promotions in case of high risk customers.

Clustering:

Instead of building a regression model for overall sales, we can cluster stores based on store type (like supercenter/ Neigborhood Market etc) and then build our models on top of it.

Instead of classifying into high and low risk customers, we first cluster customers based on income, so that like customers fall together. This will improve the accuracy of the model.