Neuro Progression is a unique tool based on powerful Neural Networks to learn any suite of numbers and predict future outputs !
Based on the NeuroNet Technology using multi-input, multi-hidden neurones and single output fully bespoke Neural Networks.
SIMPLE & EASY TO USE --------------------
- Use the efficient touch interface to define your training progression or desired output
- Define the property of your neural network
- Use an advanced genetic algorithm to train your network
- Use robust back propagation to fine tune your solution
BESPOKE NEURAL NETWORK --------------------
- Define the nature and complexity of your own network
- Choose the number of input parameters (up to 5 inputs)
- Choose the number of preceding numbers within the defined progression that have an impact on a given output
- Choose the number of hidden layers (up to 4 layers)
- Choose the number of neurones on each hidden layers (up to 20 neurones each)
- Choose the type of activation function
- Choose thresholds
FULL CONTROL --------------------
- Fine tune settings of the genetic algorithm
- Fine tune settings of the back propagation method
- Combine both genetic algorithm and back propagation for best results
- Save best settings for re-use on different problems or compare best results
EXPLORE TRAINING AND PREDICTED PROGRESSION --------------------
- Once neural network is trained, explore future predictions
- Define input to future outputs
- Visualise training and predicted progression in graphical interface
PROJECT MANAGEMENT --------------------
- An unlimited number of projects can be saved and loaded with any names
- Genetic algorithm, back propagation and network settings can be saved and recalled independently
EXAMPLES --------------------
- Neuro Progression comes with several examples
- Examples include training set, solution and settings
APPLICATIONS --------------------
Neuro Progression is a powerful that can have many engineering and financial applications
- Predict or extrapolate suite of numbers
- Learn normal system response and identify abnormal signals