mlask Namespace Reference

Classes
class	ActivationFunction
	base class for Activation Functions, meant for simplicity in definition. More...
class	ArchitectureError
	Errors associated with Neural Network architecture e.g. wrong connection beetwen layers. More...
struct	BinaryCrossEntropy
	Binary Cross-Entropy Error Function. More...
struct	DerivedBinaryCrossEntropy
	Derived Binary Cross-Entropy Error Function. More...
struct	DerivedMeanAbsolute
	Derived Mean Absolute Error Function. More...
struct	DerivedMeanSquared
	Derived Mean Squared Error Function. More...
class	ExportError
	Errors associated with ONNX export. More...
class	FullyConnectedLayer
	Class representing fully connected layer. More...
class	GenericErrorFunction
	Class representing an error function Children should define a way to move forawrd and backward, but for simplicity they define a function that operates on a single x_1 and x_2 from result vector and expected vector, instead of on the entire vector. More...
class	LambdaActivationFunction
	ActivationFunction class. More...
class	Layer
	Base class for all layers in the neural network. More...
class	LeakyRelu
	Class representing Leaky Relu activation function. More...
struct	MeanAbsolute
	Mean Absolute Error Function. More...
struct	MeanSquared
	Mean Squared Error Function. More...
class	Model
	Class representing a neural network model. More...
class	ProgressBar
	Class used to write progress bars in the terminal. More...
class	Relu
	Layer representing Relu activation function. More...
class	Sigmoid
	class reprsenting Sigmoid activation function More...
class	SoftMax
	class reprsenting Soft Max activation function, since the definition is to complex for activation function it derives from Layer More...
class	Tanh
	class reprsenting Hyperbolic Tangent activation function More...

Concepts
concept	ErrorFunction
	Concept representing an error function that contains a functor inside it @detials a class that needs to have an operator () with two float_t args, and which returns float_t(error)
concept	TLayer
	concept ensuring a class derives from Layer

Typedefs
using	float_t = float
using	size_t = std::size_t
using	int_t = std::int32_t
using	uint_t = std::uint32_t
using	vectorIn = Eigen::Matrix< float_t, Eigen::Dynamic, 1 >
using	vectorOut = Eigen::Matrix< float_t, Eigen::Dynamic, 1 >
using	err_function = float_t(*)(float_t, float_t)
using	actfunc = std::function< float_t(float_t)>

Typedef Documentation

actfunc

using mlask::actfunc = typedef std::function<float_t(float_t)>

err_function

using mlask::err_function = typedef float_t (*)(float_t, float_t)

float_t

using mlask::float_t = typedef float

All types used in this project are coded here, you can always override them in any file

int_t

using mlask::int_t = typedef std::int32_t

size_t

using mlask::size_t = typedef std::size_t

uint_t

using mlask::uint_t = typedef std::uint32_t

vectorIn

using mlask::vectorIn = typedef Eigen::Matrix<float_t, Eigen::Dynamic, 1>

vectorOut

using mlask::vectorOut = typedef Eigen::Matrix<float_t, Eigen::Dynamic, 1>