Hello Neha,

The actual weight of a variable in a simple linear composite is proportional to its standard deviation. The "B" variable has a much smaller SD than does the "A" variable, so, without modification, the "A" variable will play a much bigger role in any clustering method you use that is based on proximity quantified as distance.

1. Convert both sets to z-scores.

2. Multiply each z-converted score in the "B" set by the desired weight (relative to 1).

3. Apply clustering method of choice, based on proximity measure of choice.

Good luck with your work.

Completely agree with the answer given by David. A is having a larger SD and thus clustering is going in favor of A. However I feel that by introducing weights for both the data column, the clustering will be biased.

Use entropy based information gain.

I wanted to know how to calculate weight value for each feature in Wisconsin breast cancer dataset. So there are 30 features there should be 30 weight value. How to calculate weight value