NARX simulator with neural networks

This projects aims at creating a simulator for the NARX (Nonlinear AutoRegressive with eXogenous inputs ) architecture with neural networks. The system can fallback to MLP ( multi layer perceptron ), TDNN ( time delay neural network ), BPTT ( backpropagation through time ) and a full NARX architecture.
The system is intended to be used as a time series forecaster for educational purposes.
This projects is my personal master thesis developed at the Master of Artificial Intelligence at Universitat Polytecnica de Catalunya, Barcelona
The project presents artificial generated data but also real data tests from temperature, weather, economy.
Please refer to the documentation for test results.
If you want to improve or be part of this project, please send me a message.

You’re probably paying too much for cell phone service. Wirefly compares hundreds of plans to help you save. Enter what you need (minutes, data, texts) into Wirefly’s innovative plan comparison tools and see your savings instantly.