Wednesday, 12 September 2012

Session 9 & 10 - Group F


Session 9& 10
Multidimensional scaling and permap

Multidimensional scaling (MDS) is a series of techniques that helps the analyst to identify key dimensions underlying respondents’ evaluations of objects. It is often used in Marketing to identify key dimensions underlying customer evaluations of products, services or companies. It is based on the comparison of objects and each object has either of the two dimensions attached to it, either it objective or percieved. It helps determine:
(1) what dimensions respondents use when evaluating objects 
(2) how many dimensions they may use in a particular situation 
(3) the relative importance of each dimension
(4) how the objects are related perceptually

The purpose of MDS is to transform consumer judgement of similarity or preference into distances  represented in multidimensional space. The perceptual map shows the relative positioning of all objects. 

So, what are permaps?

It is an interactive program for making perceptual mapping. The program  uses pairwise numerical values (correlation, proximities, dissimilarities, etc) to construct a map showing the relationship between objects. A unique feature of PERMAP is that it embeds the mapping techniques in an interactive, graphical system that minimizes several difficulties associated with multidimensional scaling practices.

How does one use a PERMAP?

Understanding the screen- 

The MDS map is presented in the center of the screen.  Map identification information and the objective function value are given in the header section.

 Header and Objective Function Value
The title can be up to 65 characters long and the subtitle can be up to 98 characters long, these are also optional. This value will converge to a minimum as the iterative solution process continues.  If the Auto Repeat option is being used, then just below the current objective function value the best-found value is displayed, followed by the number of times it has been found, followed by the total number of times the problem has been solved for the given set of parameters.

Parking Active Objects Control  


The parking lots are on the right side of the main screen.  Objects placed in the object’s parking lot are removed from the active data set. This makes it easy to discover the influence of a particular object or attribute on the overall configuration.  Double click an object to move it into its parking lot.  Objects also can be dragged into the object parking lot.  To return an object to the active set just drag it out of the object parking and to the position you want it to be when the analysis continues.  Each time an object is parked or unparked, the map readjusts to represent the new active data set.  


Parking Active Attributes Control
If your data file contains attribute information, just below the functional shortcut buttons will be a box that shows the "active attributes."  These attributes are used to calculate the dissimilarities by using any of several built-in attribute-to-dissimilarity functions.  

Execution Control

A set of execution control buttons appears just below the map in the center of the screen.  Execution of the program can be started, stopped, continued, jiggled, subjected to field movements, automatically stopped, or automatically started by using these controls.

Start, Stop, Continue


When Start is clicked, the objects are placed at random positions on the map and the iterative solution procedure is commenced.  The Stop button stops the iteration and displays the object labels if the labeling option has been selected.  If the solution has converged to a stationary value of the objective function but a stop signal has not been given, then the ellipses following the objective function value continue to blink.  If you want to resume iteration after a stop command has been sent (either from the Stop button, from Auto Stop, or from a click on an open area of the map) without using new starting positions, then click the Continue button.

In our class we learnt how to operate the permap, wherein we learnt to form clusters and deriving an analysis from the same.

By:-
Beenu Sabharwal
SIBM Bangalore
14014








No comments:

Post a Comment