Index

A B C D F G I L M P R S T W
All Classes and Interfaces|All Packages|Serialized Form

A

activate(Matrix) - Method in class io.github.nearage.jnn.model.layer.Conv

activate(Matrix) - Method in class io.github.nearage.jnn.model.layer.Dense

activate(Matrix) - Method in class io.github.nearage.jnn.model.layer.Flatten

activate(Matrix) - Method in class io.github.nearage.jnn.model.Sequential

activate(Matrix) - Method in class io.github.nearage.jnn.processing.Layer

Activates the layer for the given input

activate(Matrix) - Method in class io.github.nearage.jnn.processing.Model

Generates an activation for each layer of the model

Activation - Interface in io.github.nearage.jnn.processing

Activation functions

add(Matrix, Matrix) - Static method in interface io.github.nearage.jnn.util.Matrices

Adds matrix b to matrix a

apply(Matrix) - Method in interface io.github.nearage.jnn.processing.Activation

Applies the activation function to the given input

apply(Matrix, Matrix) - Method in interface io.github.nearage.jnn.processing.Loss

Applies the loss function to the given input

apply(Function<Double, Double>) - Method in class io.github.nearage.jnn.input.Matrix

Sets the each value in the matrix to the result of applying the given function to that value

avg() - Method in class io.github.nearage.jnn.input.Matrix

Gets the average value of the matrix

B

biases - Variable in class io.github.nearage.jnn.processing.Layer

Biases of the layer

C

cols - Variable in class io.github.nearage.jnn.input.Matrix

Number of cols

Conv - Class in io.github.nearage.jnn.model.layer

TODO: Convolutional layer

Conv() - Constructor for class io.github.nearage.jnn.model.layer.Conv

Initializes a convolutional layer

D

Dataset - Class in io.github.nearage.jnn.input

TODO: Dataset

Dataset(Matrix[], Matrix[]) - Constructor for class io.github.nearage.jnn.input.Dataset

Initializes a new dataset with the given batches

Dense - Class in io.github.nearage.jnn.model.layer

Dense layer

Dense(int, Activation[]) - Constructor for class io.github.nearage.jnn.model.layer.Dense

Creates a dense layer with the specified neurs and activation function

describe() - Method in class io.github.nearage.jnn.input.Matrix

Prints a description of the matrix

dot(Matrix, Matrix) - Static method in interface io.github.nearage.jnn.util.Matrices

Performs the dot product of two matrices

F

Flatten - Class in io.github.nearage.jnn.model.layer

TODO: Flatten layer

Flatten() - Constructor for class io.github.nearage.jnn.model.layer.Flatten

Initializes a flatten layer

foreach(Consumer<Double>) - Method in class io.github.nearage.jnn.input.Matrix

Performs an action for each element in the matrix

G

generate_sum(int, int, int) - Static method in class io.github.nearage.jnn.input.Dataset

Generates a demonstration dataset as a random matrix of the specified shape, and a target matrix that classifies the values of the sum of each input.

get(int, int) - Method in class io.github.nearage.jnn.input.Matrix

Gets the value at the given index

I

inputs - Variable in class io.github.nearage.jnn.input.Dataset

Input batches

io.github.nearage.jnn.input - package io.github.nearage.jnn.input

io.github.nearage.jnn.model - package io.github.nearage.jnn.model

io.github.nearage.jnn.model.layer - package io.github.nearage.jnn.model.layer

io.github.nearage.jnn.processing - package io.github.nearage.jnn.processing

io.github.nearage.jnn.util - package io.github.nearage.jnn.util

iterate(BiConsumer<Integer, Integer>) - Method in class io.github.nearage.jnn.input.Matrix

Performs an action for each index in the matrix

L

Layer - Class in io.github.nearage.jnn.processing

Layer

Layer() - Constructor for class io.github.nearage.jnn.processing.Layer

Initializes a layer

layers - Variable in class io.github.nearage.jnn.processing.Model

Layers of the model

load(String) - Static method in class io.github.nearage.jnn.input.Matrix

Loads a matrix from a file in the specified path

Loss - Interface in io.github.nearage.jnn.processing

Loss functions

M

map(BiFunction<Integer, Integer, Double>) - Method in class io.github.nearage.jnn.input.Matrix

Sets the value for each index in the matrix to the result of applying the given function to that index

Matrices - Interface in io.github.nearage.jnn.util

Matrix utils

Matrix - Class in io.github.nearage.jnn.input

Matrix of doubles

Matrix(int, int) - Constructor for class io.github.nearage.jnn.input.Matrix

Creates a new matrix with the specified shape

max() - Method in class io.github.nearage.jnn.input.Matrix

Gets the max value in the matrix

MeanSquaredError - Static variable in interface io.github.nearage.jnn.processing.Loss

Mean Squared Error loss function

min() - Method in class io.github.nearage.jnn.input.Matrix

Gets the min value in the matrix

Model - Class in io.github.nearage.jnn.processing

Model

Model(Layer...) - Constructor for class io.github.nearage.jnn.processing.Model

Creates a model with the given layers

mul(Matrix, Matrix) - Static method in interface io.github.nearage.jnn.util.Matrices

Multyplies matrix a by matrix b

P

peek() - Method in class io.github.nearage.jnn.input.Matrix

Gets the value of the first element

predict(Matrix) - Method in class io.github.nearage.jnn.model.Sequential

predict(Matrix) - Method in class io.github.nearage.jnn.processing.Model

Generates output predictions for the input samples

print() - Method in class io.github.nearage.jnn.input.Matrix

Prints every value of the matrix

propagate(Matrix) - Method in class io.github.nearage.jnn.model.layer.Conv

propagate(Matrix) - Method in class io.github.nearage.jnn.model.layer.Dense

propagate(Matrix) - Method in class io.github.nearage.jnn.model.layer.Flatten

propagate(Matrix) - Method in class io.github.nearage.jnn.processing.Layer

Propagates the activation of the layer for the given input

R

randomize(int, int) - Method in class io.github.nearage.jnn.input.Matrix

Randomizes the elements in the matrix within the given bounds, both included

RectifiedLinearUnit - Static variable in interface io.github.nearage.jnn.processing.Activation

Rectified Linear Unit activation function

reduce(double, BiFunction<Double, Double, Double>) - Method in class io.github.nearage.jnn.input.Matrix

Reduces the matrix to a single value, applying the given identity and function

reduce(int, double, BiFunction<Double, Double, Double>) - Method in class io.github.nearage.jnn.input.Matrix

Reduces the matrix along the specified axis, applying the given identity and function

rows - Variable in class io.github.nearage.jnn.input.Matrix

Number of rows

S

save(String) - Method in class io.github.nearage.jnn.input.Matrix

Saves the matrix to a file in the specified path

Sequential - Class in io.github.nearage.jnn.model

Sequential model

Sequential(Layer...) - Constructor for class io.github.nearage.jnn.model.Sequential

Creates a Sequential model with the given layers

set(int, int, double) - Method in class io.github.nearage.jnn.input.Matrix

Sets the given index to the given value

Sigmoid - Static variable in interface io.github.nearage.jnn.processing.Activation

Sigmoid activation function

size - Variable in class io.github.nearage.jnn.input.Matrix

Size of the matrix

Softmax - Static variable in interface io.github.nearage.jnn.processing.Activation

Softmax activation function TODO: Softmax propagation

sub(Matrix, Matrix) - Static method in interface io.github.nearage.jnn.util.Matrices

Substracts matrix b to matrix a

sum() - Method in class io.github.nearage.jnn.input.Matrix

Gets the sum of all values in the matrix

summary() - Method in class io.github.nearage.jnn.input.Matrix

Prints a summary of the matrix

summary() - Method in class io.github.nearage.jnn.model.Sequential

summary() - Method in class io.github.nearage.jnn.processing.Model

Prints a string summary of the model

T

targets - Variable in class io.github.nearage.jnn.input.Dataset

Target batches

train(int, Matrix, Matrix, Loss[], double) - Method in class io.github.nearage.jnn.model.Sequential

train(int, Matrix, Matrix, Loss[], double) - Method in class io.github.nearage.jnn.processing.Model

Trains the model for a fixed number of epochs

transpose() - Method in class io.github.nearage.jnn.input.Matrix

Transposes the matrix

W

weights - Variable in class io.github.nearage.jnn.processing.Layer

Weights of the layer

A B C D F G I L M P R S T W
All Classes and Interfaces|All Packages|Serialized Form