tiny_dnn::adagrad Struct Reference

adaptive gradient method More...

#include <optimizer.h>

Inheritance diagram for tiny_dnn::adagrad:

Collaboration diagram for tiny_dnn::adagrad:

Public Member Functions
void	update (const vec_t &dW, vec_t &W, bool parallelize)
Public Member Functions inherited from tiny_dnn::stateful_optimizer< 1 >
void	reset () override
Public Member Functions inherited from tiny_dnn::optimizer
	optimizer (const optimizer &)=default
	optimizer (optimizer &&)=default
optimizer &	operator= (const optimizer &)=default
optimizer &	operator= (optimizer &&)=default

Public Attributes
float_t	alpha

Additional Inherited Members
Protected Member Functions inherited from tiny_dnn::stateful_optimizer< 1 >
vec_t &	get (const vec_t &key)
Protected Attributes inherited from tiny_dnn::stateful_optimizer< 1 >
std::unordered_map< const vec_t *, vec_t >	E_ [N]

Detailed Description

adaptive gradient method

J Duchi, E Hazan and Y Singer, Adaptive subgradient methods for online learning and stochastic optimization The Journal of Machine Learning Research, pages 2121-2159, 2011.

Member Function Documentation

update()

void tiny_dnn::adagrad::update ( const vec_t &  dW, vec_t &  W, bool  parallelize  )

inlinevirtual

Implements tiny_dnn::optimizer.

---

The documentation for this struct was generated from the following file:

• tiny_dnn/optimizers/optimizer.h