Cluster analysis techniques or “clustering” form one of the three most commonly used data science algorithms. A very popular use of cluster analysis is for market segmentation where customers are segmented based on demographics and transaction data. The goal is to formulate a market strategy for each segment.
Statsoft, the makers of STATISTICA data mining software, have released a set of 35 video tutorials which are a collection of data mining techniques that their software is capable of. Cluster Analysis is described in Session 28 of this video series.
What it is: A 9-min introduction to running a cluster analysis using their software. However the video prior to this, which is a 8-min Session 27 – Introduction to Marketing Data, must be viewed first to obtain the full picture of the project objective and how cluster analysis is used to understand the data.
The great part: Very good and quick introduction to running a k-means cluster analysis using their tool. Starting with session 27, the data prep video, one can very easily set up a new analysis and potentially start generating insights into data.
Useful Shortcuts: “What is cluster analysis” 0:34, “Options available for cluster analysis techniques” 1:08
What could be better: Very cursory distinction made between when the different cluster analysis techniques are appropriate. Also if you are looking for any information into how clustering actually works, this video is not for you. The full value only comes when you use this in combination with their software.
Was this review helpful?
Originally posted on Mon, Sep 26, 2011 @ 11:08 AM