In the "Model Building" step of CRISP$-$DM$,$ why are multiple modeling techniques explored?

Question $26$ options:

a$)$

To ensure that data preparation is adequately tested.

b$)$

Different techniques may reveal different patterns in the data.

c$)$

Because all data mining tools automatically select the best model.

d$)$

It is a regulatory requirement in most data mining projects.

e$)$

Only one technique is necessary but multiple are used for experimentation.