Released in the AssetStore:
A set of scripts for creating and learning your own artificial neural network.
"ANN Perceptron" is also present and supported.
Creating your own artificial neural network (ANN) is not a big issue. But to teach the ANN to do certain tasks is a real challenge. In order to ease the task of creating and teaching the ANN, I wrote a set of scripts, that will do everything for you. The only thing that is needed from you is to “explain” the ANN properly, what information is needed to be learned.
Flexible settings for the artificial neural network and its learning.
Two methods of learning for "ANN Perceptron":
- Back propagation.
- Random Generation.
One methods of learning for "ANN":
- NEAT (NeuroEvolution of Augmenting Topologies).
The auxiliary interfaces.
Tutorial in four lessons.
Sergey Voroshilov Vladimirovich
VirtualSUN - Programmer
Ver. 1.066: - Fixed signatures in scripts. - Fixed update visualization of the activation function in the interfaces. - Correction in the instructions. - Added ability to save / load learning settings. - Added demonstration "Car2D". - Perceptron update: - In PerceptronLernByBackPropagation, “LearningSpeed” is limited to 1 second to avoid freezes. "LearningSpeed" ​​continues to work in the next frame. - Fixed errors associated with the creation of a perceptron by scripting. - Minor changes in the perceptron interface. - ANN update: - Added the ability to add "memory connections" between neurons (except for input neurons) during training. - The neurons of the hidden layer can now get the connection between themselves in the form of "memory connections". - Added the ability to obtain "memory" from neurons. - Removed "Bias" in ANN.cs and ANNInterface.cs as unnecessary. Now the presence of Bias in neurons will depend on learning. - Additions in the visualization of weights. - Minor ANN solution optimization. - Reworked algorithm for calculating possible links. - Add-on interfaces related to NEAT. - Added "TennisANN" demo for ANN. - Added the ability for neurons to use different activation functions during training. - Fixed ANN solution after loading from the perceptron file.
Ver. 1.019: - Added new activation functions ("ReLU", "Logistic", "TanH", "Sinusoid", "Sinc", "Gaussian"). - Added visualization of activation functions. - Activation functions are moved to "ActivationFunctions.cs". - Minor interface fixes. - Fixed the error in determining the type of activation function for the perceptron. - Fixed calculation of derivatives for all formulas for the backpropagation method ("Identity" and "Binary step" are temporarily not suitable for this training method). - Calculation of derivatives moved to "ActivationFunctions.cs". - Minor bugfix for backpropagation method. - Added message about incorrect input data. - Fixed visualization of weights. - Now the instruction is in one file. - Internal and external links are created in the instruction.
Updated to ver. 1.006: - The main folder was renamed from "ANN&TOOLS to "ANN" (for better compatibility with "ANN perceptron" and "ANN & NEAT"). - New activation functions. - The calculation of the best generation was fixed. - Fixed number of generation for new training. - Fixed visualization of neuron connections. - The activation function migrated from ANN.cs and Perceptron.cs to Formulas.cs.