Aggregated Flexibility Forecast Using Online Recursive Maximum Likelihood Kernel Density Estimation - Associated Data and Code

Published: 31 March 2021| Version 1 | DOI: 10.17632/t92mmtm4gs.1
Contributors:
Ingrid Munné-Collado,
Pierre Pinson

Description

This data repository contains the associated code to reproduce the formulation regarding Online Recursive Maximum Likelihood Kernel Density Estimation. The input data are measurements, while the output data are the density function updated recursively when a new measurement enters the model. Accordingly, auxiliary functions to compute the log-likelihood performance score as well as safety checks and results plots functions are included. This approach recursively estimates the kernel bandwidth at each time step or iteration, in order to maximize the likelihood. The formulation on which these scripts are based can be found in: "Aggregated Flexibility Forecast Using Online Recursive Maximum Likelihood Kernel Density Estimation" by Munné-Collado, Íngrid; Pinson, Pierre; Aragüés-Peñalba, Mònica and Sumper, Andreas; submitted to Applied Energy.

Files

Steps to reproduce

The README.md file includes all information and steps to reproduce this research.

Institutions

Universitat Politecnica de Catalunya, Danmarks Tekniske Universitet

Categories

Forecasting, Online Learning, Kernel Density Estimation, Flexibility, Maximum Likelihood

License