[source]

Xavier 2

The Xavier 2 initializer is a backward-compatible alias of He. Like He, it draws from a uniform distribution with limits of +/- sqrt(6 / fanIn). It is kept to preserve the name for existing configurations.

Note

Xavier 2 is deprecated, use He instead.

Parameters

This initializer does not have any parameters.

Example

use Rubix\ML\NeuralNet\Initializers\Xavier2;

$initializer = new Xavier2();

References

---

1. X. Glorot et al. (2010). Understanding the Difficulty of Training Deep Feedforward Neural Networks. ↩