Question

For each of the following scenarios, decide if a solution would be best addressed with supervised learning, unsupervised learning or database query. As appropriate, state initial hypotheses you would like to test. If you decide that supervised or unsupervised is the best answer, list several input attributes you believe to be relevant for solving the problem.

a. What characteristics differentiate people who have had back surgery and have returned to work form those who have had back surgery and have not returned to their jobs?

b. A major automotive manufacturer recently initiated a tire recall for one of their top-selling vehicle. The automotive company blames the tires for their unusually high accident rate with their top-seller. The company producing the tires claims the high accident rate occurs only when their tires are on the vehicle in question. Who is to blame?

c. When customers visit my website, what products are they most likely to buy together?

d. What percent of employees miss one or more days of work per month?

e. What relationships can be found between an individual’s height, weight, age, and favorite spectator sport?

Answer 1

a] Supervised Learning: In order to know what characteristics differentiate people, we need to know all the characteristics of the people who has back surgery and the data about whether they have returned to work or not. based on this we build a model and see the features/characteristics which have more weight in model, or which play crucial role in classification.

b] Database Query: We need to know whom to blame? For that we go through the data and get all the data when that tire is used. now, see the vehicle in those data. If all the vehicles belong to automotive company then it's automotive company's fault. If there are other vehicles of other companies then it's tyre manufacturer's fault.

c] Unsupervised Learning: In order to know what products the customer most likely buy together, we need to know what products the customer brought together in the past. and based on that we will cluster the products and tell from those clusters what is the user likely to buy together?

d] Database Query: We can just go through the database and see the number of employees who have missed one or more days of work and calculate the percentage.

e] Supervised Learning: We need to know the data of height, weight, age and favourite spectator sport of some persons and build the linear or non linear model based on data and see the weights of the features to find the relationship between them.