Analytics is the most sophisticated analysis of data.

Analytics closely resembles statistical analysis and data mining, but tends to be based on physics modeling with extensive computation.

Analytics can progress along two lines:

1. Using existing patterns/metrics/performance indicators (KPIs):
   1. Industry generic metrics
   2. Company specific metrics
2. Using Sophisticated statistical/datamining tools to derive new patterns/metrics/performance indicators.

Banking example

A common application is portfolio analysis. In this, a bank or lending agency has a collection of accounts, some from wealthy people, some from middle class people, and some from poor people. The question is how to evaluate the whole portfolio.

The bank can make money by lending to wealthy people, but there are only so many wealthy people. The bank can make more money by also lending to middle class people. The bank can make even more money by lending to poor people.

Note that poorer people are usually at greater risk of default. Note too, that some poor people are excellent borrowers. Note too, that a few poor people may eventually become rich, and will reward the bank for loyalty.

The bank wants to maximize its income, while minimizing its risk, which makes the portfolio hard to understand.

The analytics solution may combine time series analysis, with many other issues.

See also

List of software engineering topics, Statistics