A neural network is a computational model inspired by the structure and function of the human brain. It consists of layers of interconnected nodes, or "neurons," that process data and learn patterns through training.
With NeuralNet Pro you can implement neural networks in your game or apps, without additional installation, everything is included in one package.
Imagine NeuralNet as a Function, you give an Input, and you’ll receive an Output. It’s that simple, but in pure NeuralNetwork you need to convert the input into an array of float, between 0-1. and the output is also just an array of float 0-1.
If you are interested to know how to use Pure NeuralNetwork, please read the article below.
Pure Neural Network, without additional helper component
Good News, This NeuralNet Pro has a life-saving helper to make you run NeuralNet easier, NeuralInput that converts game parameters to arrays required by NeuralNetwork, and NeuralOutput that immediately provides you with Continous and Discrete results.
Get back to how to make it run, you need these steps to run a NeuralNet
Firstly, you need to feed a NeuralNet sufficient information to provide output you wanted. that’s why you need to set up your Input and Output.
Secondly, to run a NeuralNet, you’ll need a Neural Memory, which is a model you get after training is completed. Goto Training
Machine learning training is divided into 2, Supervised and Unsupervised Learning. it can be different from each case, you need to choose your learning method depending on the situation.
For training a model, you also need to understand the Hyperparameter and its function to train the model properly and efficiently.
Hyper Parameter