18 lines
510 B
Matlab
18 lines
510 B
Matlab
function [X_norm, mu, sigma] = featureNormalize(X)
|
|
%FEATURENORMALIZE Normalizes the features in X
|
|
% FEATURENORMALIZE(X) returns a normalized version of X where
|
|
% the mean value of each feature is 0 and the standard deviation
|
|
% is 1. This is often a good preprocessing step to do when
|
|
% working with learning algorithms.
|
|
|
|
mu = mean(X);
|
|
X_norm = bsxfun(@minus, X, mu);
|
|
|
|
sigma = std(X_norm);
|
|
X_norm = bsxfun(@rdivide, X_norm, sigma);
|
|
|
|
|
|
% ============================================================
|
|
|
|
end
|