I am trying to identify the most similar stations between two DataFrames like below:

stations      feature_1     feature_2   feature_3 ------ feature_10
08GD008         10           1.14          98
08GE002          5           88.67         80
08MC040          8           4.61          17
08FB006          2           13.70         53       
08FC003          1           37            49
08LF002         20           2.5           30

I used a pairwise Euclidean distance and identified the minimum distance for each station to select the candidate (most similar one). I used (x-mean)/std to standardize my features ( they all have equal weights). I wanted to check if there is a way to improve my method.

I thought of using PCA (principal component analysis) to use PC1 and PC2 for my distance matrix but only 60% of variability is explained within the first two components. I was wondering if there is a way to assign weights to my features ( something like coefficient of variation of each feature multiplied by all standardized values)? or something similar to PCA in this case.


migrated from stackoverflow.com Apr 2 at 22:31

This question came from our site for professional and enthusiast programmers.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.