BCG Interview Question | Trial & Error
Question
What is the role of trial and error in data analysis? Is making a hypothesis before diving in important?
in progress
0
Statistics
4 years
2 Answers
1085 views
Great Grand Master 0
Answers ( 2 )
A hypothesis is very much necessary before diving into conclusion and causality.
When we explore data, it will tell us many things and some of them can be misleading till we test them statistically by forming hypothesis and check the significance level.
In data analysis, trial and error helps in unearthing the additional evidence which would
help us to make our estimates better.
Also while building ML models, we try to do feature engineering for increasing model’s accuracy,
which is a kind of trial and error because we don’t know whether it will really help in doing that.
Hypothesis is important before making any kind of analysis because it gives a clear purpose to the
kind of analysis you are doing