A closest neighbor can be a row of data (a case) that's just like a concentrate on case. The gap among any two cases is measured by combining the weighted vectors of all features. . There are lots of explanation why a dataset may be imbalanced. By way of example, https://harvard-case-study-soluti66805.weblogco.com/36101782/the-smart-trick-of-online-case-solution-that-no-one-is-discussing