documentation: nearest neighbors

17 次查看（过去 30 天）
Rich Tans
Rich Tans 2018-4-30
回答： Walter Roberson 2018-4-30
Classification using NN: Distance metric definition equations. Impossible to make out exponent of (xs-yt)^? term for Euclidean distance, same for Standard Euclidean and same for Mahalanobis ( Xs - Yt)^ ?, Similar problems with Cosine distance and Correlation distance. I tried zooming to 300 % , still ambiguous.I know what textbooks say these should be and I can check that but wanted to see how Matlab is treating these terms. Suggest you either make these coeff larger, or call then a variable ( like you did for p in other definitions) .

请先登录，再回答此问题。

回答（1 个）

Walter Roberson
Walter Roberson 2018-4-30
If you are referring to https://www.mathworks.com/help/stats/classification-using-nearest-neighbors.html then there is no exponent for the Euclidean distance given there. The mark given in the documentation is the ' transpose operator.

类别

在 Help Center 和 File Exchange 中查找有关 Nearest Neighbors 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by