Tuesday, September 11, 2012

Day 5 Team C (Nitin)


Team C
By-(NitinDhantole)
BA as on 11-9-12
MDS- Multi dimensional scale.
Permap- perceptual mapping.
1.       Overall similarity- done for each pair of object to get similarity for the pair. It is more flexible. We find out overall similarity between the attributes in the diagram which can be known by knowledge abt the attributes. Advantage is that it gives which is not on the mind like the black box of the mind attributes. Disadvantage is to have thorough knowledge abt the attributes
2.       Attribute based- Rating of attributes is given by the people and then the rating is given is attitude based permap this is based on the knowledge of the other people. Advantage is simpler compared to similarity. It is done on the scale of the attributes and rating is given to the people. It develops something which is not done by our side in rating scale

Took example of soft drinks in excel and pairs were given rating in between 0-9 and given higher rating that is similar pairs. In this we take average of the pair rating and here higher the average higher the similarity. In the distance matrix 1 and 1 represent higher similarity. It can be a similarity matrix or dissimilarity matrix in which the 1 or 0 in the diagonal represent the matrix. Then copy the matrix in row and save it in note pad and load the file in permap software. There are minor differences in the map created in permap it is due to different laptops and there configurations.

In permap to check the minimum distance in the diagram we use auto repeat and auto stop.
By use of hierarchical clustering we can develop matrix on proximity matrix and can use it in permap software
Park objects in permap leads to remove a variable form the mapped data. By this mapping we can segregate the map and the perception in each dimension. Whenever the value drops substantiallyit is 3 dimensional. 

No comments:

Post a Comment