README
¶
cmplxNN
golang implementation of nuetral net using complex numbers
Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
View Source
var Normalize = func(c complex128) complex128 { abs := cmplx.Abs(c) return c * complex(1/abs, 0) * complex(-math.Expm1(-abs), 0) }
Normalize is a function for normalizing the output values of Nueral Network Layers. is a variable that can be replaced with another function to allow to alternate Normalization functions
Functions ¶
This section is empty.
Types ¶
type Layer ¶
type Layer struct {
// Weights is a matrix of complex128 weights, w[row][column] indexing.
// the number of columns is the expected dimension of the input array
// the number of rows is the dimension of the output array
Weights [][]complex128
// Bias is added to the output after adjusting based on the weight and before normalizing the values.
// the length of the bias should match the number of rows in Weights
Bias []complex128
}
Layer is a matrix representation of a layer of a nueral net.
func (*Layer) Generate ¶
func (l *Layer) Generate(in []complex128) (out []complex128)
type Model ¶
type Model interface {
Generate([]complex128) []complex128
}
type NueralNet ¶
type NueralNet struct {
Layers []Layer
}
func (*NueralNet) Generate ¶
func (nn *NueralNet) Generate(in []complex128) (out []complex128)
Source Files
Directories
Click to show internal directories.
Click to hide internal directories.