hgalib-0.2: Haskell Genetic Algorithm Library Source code Contents Index

Chromosome.ANN

Description

Artificial Neural Networks

Synopsis

type ANN = [Layer] type Layer = [Node] type Node = [Double] eval :: [Double] -> ANN -> [Double] config uniformCross :: ANN -> ANN -> GAState c p (ANN, ANN) averageCross :: ANN -> ANN -> GAState c p (ANN, ANN) mutateRandomize :: Double -> Double -> ANN -> GAState ANN p ANN mutateShift :: Double -> Double -> ANN -> GAState ANN p ANN fitnessMSE :: [([Double], [Double])] -> ANN -> Double averageMSE :: ANN -> [([Double], [Double])] -> Double correctExamples :: [([Double], [Double])] -> Double -> ANN -> Double randomANN :: Int -> [Int] -> Double -> GAState c p ANN

Documentation

type ANN = [Layer] Source

An Artificial Neural Network

type Layer = [Node] Source

A layer of nodes in an ANN

type Node = [Double] Source

A node in an ANN. The head of the list is the bias weight. The tail is the weights for the previous layer

eval :: [Double] -> ANN -> [Double] Source

Evaluates an ANN with a given input

config

uniformCross :: ANN -> ANN -> GAState c p (ANN, ANN) Source

Crossover between two ANN's by exchanging weights

averageCross :: ANN -> ANN -> GAState c p (ANN, ANN) Source

Crossover between two ANN's by averaging weights

mutateRandomize :: Double -> Double -> ANN -> GAState ANN p ANN Source

Mutates an ANN by randomly settings weights to +/- range

mutateShift :: Double -> Double -> ANN -> GAState ANN p ANN Source

Mutates an ANN by randomly shifting weights by +/- range

fitnessMSE :: [([Double], [Double])] -> ANN -> Double Source

Computes the fitness based on the mean square error for a list of examples The examples are a list of tuples of (input, output)

averageMSE :: ANN -> [([Double], [Double])] -> Double Source

Computes the mean square error for a list of examples The examples are a list of tuples of (input, output)

correctExamples :: [([Double], [Double])] -> Double -> ANN -> Double Source

Returns the number of examples correct within the tolerance. The examples are a list of tuples of (input, output)

randomANN :: Int -> [Int] -> Double -> GAState c p ANN Source

Generates a random ANN with a given number of input nodes, a list of number of hidden nodes per layer, and the weight range

Produced by Haddock version 2.4.2