narx

A Python Package For System Identification Using NARMAX Models

A python multi-variate time series prediction library working with sklearn

The project uses a nonlinear autoregressive exogenous (NARX), model to make time-series prediction on data obtained from drive cycling testing on buses

A fast canonical-correlation-based search algorithm for feature selection, system identification, data pruning, etc.

Air-quality forecasting in Belgium using Deep Neural Networks, Neuroevolution and distributed Island T​ranspeciation

Deep Learning using Neural Network Toolbox + Finance Portfolio Selection with MorningStar

Explore the double-descent phenomena in the context of system identification. Companion code to the paper (https://arxiv.org/abs/2012.06341):

GUI for System Identification using NARX and NARMAX models

A MatLab package for System Identification using linear and nonlinear auto-regresive models (N)AR, (N)ARX and (N)ARMAX models

Code, figures, animations for a NARX-EFE based agent.

Performed a nonlinear model identification using both the FROE method with polynomial NARX models and feedforward neural networks.

NARX Pytorch implementation

This repository contains code for energy forecasting using multilayer neural networks (MLPs) with autoregressive (AR) and nonlinear autoregressive exogenous (NARX) approaches. The goal is to predict the next-day electricity consumption for a specific case study, utilizing real dataset from a real-life organization.

An active inference agent based on expected free energy minimization with a nonlinear autoregressive exogenous model.

Neural Network Lecture Projects.

Nonlinear autoregressive exogenous model library. NarxSim ported to U++.

Dual-Stage Attention-Based Recurrent Neural Net for Time Series Prediction

Implementation of Robotic Arm Control from RNN and use of NARX with Evolutionary Algorithms for identification, for the Intelligent Automation class.

Developed a time series forecasting system using LightGBM and NARX models to optimize food demand prediction. Used real-world data with preprocessing techniques like normalization and lag feature generation. The proposed NARX model outperformed LGBM, improving accuracy and helping reduce food waste through better supply chain decisions.