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

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.