My previous blog, on Business Intelligence covered Descriptive Analytics, and the elements of understanding the data.
In this blog, the final part of the BI/Descriptive Analytics topic, I will cover the importance of visualisation, and why it matters for the analytics programs.
Why visualisation is important?
Visualisation benefits the end consumer in many ways, important ones being:
- Representing the underlying data in an uncomplicated way to consume
- Highlight the key facts, and be easily remembered
- Provides a clear view of their business at a glance, and drives next set of business actions
In the context of analytics, visualisation takes a higher place on the pedestal, due to the complexity of the underlying wiring. Many analytics programs have failed, or haven’t met their objectives, because the end consumer wasn’t able to “digest” the complex technical information, and eventually relate to their work.
In the next section, let’s look at commonly used charts and visualisation techniques in descriptive analytics.
|Histogram||Histograms are created for numerical data. They are used to assess the distribution of numerical data||Daily collections, Weekly sales etc|
|Bar chart||Bar charts are created for categorical data (where histograms cannot be used as the data is non- numeric). They are used to assess the most-occurring and least-occurring categories within a data set||Sales by customer category|
|Pie Chart||Pie charts are mainly used for categorical data, and used to assess the proportion of each category in the data set||Sales by customer category represented as %|
|Scatter Plot||Scatter plot is a plot of typically 2 variables, to understand the relationship between them.||Marketing spend vs sales – is there any relation between them|
|Box Plot or Whisker Plot||Box plot is used on numerical data, and used to understand the variability of the data set, and the existence of outliers. It is constructed using IQR, min and max values||Share prices represented as a daily movement – is a variant of a box plot|
|Treemap||Treemap is a hierarchical map, using nested rectangles. The size and colour of the rectangle are used to distinguish the data.||Revenue by various lines of business|
Today, some of the business intelligence tools, brings a whole variety of charts, in addition to the common ones above. Understanding the purpose and a chart, and then interpreting the data from those charts, will make the life of a user easier.
This bring us to the end of Business Intelligence & Descriptive Analytics. In the upcoming blogs, let us start exploring the concepts of Machine Learning through analytics.