NormalizeTransform<T>
Models & Types · AiDotNet.Data.Transforms.Numeric
Normalizes an array of values using mean and standard deviation: (x - mean) / std.
Implements: ITransform<T[], T[]>
For Beginners
Normalization makes your data have zero mean and unit variance, which helps neural networks train faster and more stably.
How It Works
This is the standard normalization used in image preprocessing and feature scaling. Each element is independently normalized using the provided per-channel or global mean/std.
Constructors
| Constructor | Summary |
|---|---|
NormalizeTransform(T mean, T std, int length) |
Creates a normalize transform with a single global mean and standard deviation. |
NormalizeTransform(T[] mean, T[] std) |
Creates a normalize transform with per-element mean and standard deviation. |
Methods
| Method | Summary |
|---|---|
T[] Apply(T[] input) |