What is considered normal for data science projects?
a) They are linear and follow standard procedures.
b) Problem definition is the least significant effort.
c) They are based on statistics.
d) They usually require innovation as well as trial and error.