A Neural Network Library made in Kotlin, from scratch.
Create instance of class NeuralNetwork: var nn: NeuralNetwork = NeuralNetwork(learningRate, ActivationFunction, layers)
ActivationFunctions supported: TanH, Sigmoid, Relu, Linear
All inputs and outputs are Double Arrays
For training use:
nn.trainSingle(dataset) or
nn.trainBatch(batchSize, dataset)
A dataset can be generated by using the addLabel() function or by writing a class that implements datasetGenerator