Answer to Question #217455 in Marketing for saleemaslam

Question

Question 2

Suppose you work for the university and have access to student, class, professor, department, and grade data. Assume the student data includes students’ home address, high school, and prior post-secondary performance (if any).

a)       Describe an unsupervised data-mining technique that could be used to predict which new applicants are likely to succeed academically.

b)       Is it responsible or irresponsible to use an unsupervised technique for such a problem? Briefly explain your answer.

Answer 1

a) Clustered data mining method could be used in this case to group students and determine their chances of succeeding for new applicants. The method classifies students based on similarities among cluster members. The characteristics of members from a cluster are identical in greater sense than the characteristics between members of different clusters. In this case, the data of new students could be clustered based on their origins to determine their likelihood of them succeeding academically.

b) It is not unethical to use unsupervised technique in this case since it may capture some meaningful attributes of the data that are not captured using the supervised data mining processes.