Why Time Series Analysis is so Useful

Time series analysis is a powerful analytical tool. How things change with time is highly common form of information visualization or data analysis. You see time series graphs nearly everyday in the newspapers or on the television news.

In business intelligence it is an essential. For example, are profits increasing or decreasing over time? The variable of time gives us a basis of understanding data.

The present can only be understood and the future can only be predicted in the light of the past according to Stephen Few in his book 'Now you See it' about simple visualization techniques for quantitative analysis.

Time Series Patterns

There are six basic patterns for analysing data through time and why they are so useful.

  • Trend - the series of points or values can go up, down or stay the same or flat.
  • Variability - how much the data points vary from point to point
  • Rate of Change - how big is the relative change between the points or values
  • Co-variation - what is the relationship between two variables or measures in a series either immediately or later
  • Cycles - are there patterns that repeat at regular intervals e.g. at the end of each quater when sales people close sales to hit their targets to win bonuses
  • Exceptions or Outliers - what values in a series of points fall outside the norm say when there has been some significant change - shortage, war etc
These patterns above give you a framework for analysing data through time. Look for these patterns and then you will be able to interpret better what is going on now and what might be likely to happen in the future.

Go to Top of Time Series Analysis page

Back to 'Business Intelligence' Home Page