Welcome

Login

CLUSTERING - NATURAL CULMINATION OF PARETO ANALYSIS

Tanmoy Mukhopadhyay
MBB, ANZ

One of the basic concepts what we learn in data science is Pareto principle or 80-20 rule. It says: 80% of the effects can be traced back to as few as with 20% of all causes – they (these 20% causes) are the “vital few”; and rest 20% of effects is then naturally distributed to be mapped with remaining 80% causes - they are “trivial many”. This empirical rule is called ‘Pareto principle' which essentially talks about the skewed distribution at various fields of study. Vilfredo Pareto - an Italian economist first identified this principle for skewed distribution of wealth – only 20% of population owns 80% of wealth, or may be even more skewed.

A deeper look at Pareto principle: If you carefully observe, you will see what this rule does is “empirical classification”. The name “rich” “poor” is only decided based on their wealth. So it is based on single variable –wealth. &

Read full article