4 ms·
At least one useful take-away: k-modes clustering. I initially thought it was a typo/miscommunication, but the soybean reference seemed bizarrely specific. It
by llasram 13y ago
At least one useful take-away: k-modes clustering. I initially thought it was a typo/miscommunication, but the soybean reference seemed bizarrely specific. It apparently actually exists, and is an extension of k-means to categorical data, using Hamming distance instead of Euclidean distance.
- wjnc 13y agoGood answer on stats.stackexchange [1] regarding k-means, standardization and discrete or categorical data [1] http://stats.stackexchange.com/a/58920 http://stats.stackexchange.com/a/58920