Question

Outline the best clustering method for the following tasks (and briefly describe the reason you would make such a design) :

(a) Finding oil spills along a coast line.

(b) Clustering employees in a company based on their salaries and years of working experience.

Answer 1

(b)

clustering employees in a company based on their salaries and years of working experience is pretty straight forward.

Few partitions need to to be done , clustering is all about dividing the data into partitions so that retreiving of data can be made faster than normal storage of data .

the partitions for this scenario are salary and experience .

salary is proprtional to experience in any company .so first partition the experience into few stages . 0-2 years , 2-5 years , 5-8 years , 8-12 years .12+ years .

based on this partition . one can divide the salaries into the respective partition .

0-2 years can contain all the employees salaries ranging from 3 - 5 .

2-5 years can contain all the employees salaries ranging from 5-7

5-8 years can contain all the employees salaries ranging 8-12.

8-12 years can contain all the employees salaries ranging 12-15 .

12+ years can contain all the employees salaries ranging 15+

hence u place all the employees under the partition so that searching can be done very fast