arXiv subjects

Francisco Martínez

Publications and source records attributed to Francisco Martínez.

4 recordsLinked to original sources

arXiv2026

Automated univariate time series forecasting with regression trees

This paper describes a methodology for automated univariate time series forecasting using regression trees and their ensembles: bagging and random forests. The key aspects that are addressed are: the use of an autoregressive approach and recursive forecasts, how to select the autoregressive features, how to deal with trending series and how to cope with seasonal behavior. Experimental results show a forecast accuracy comparable with well-established statistical models such as exponential smoothing or ARIMA. Furthermore, a publicly available software implementing all the proposed strategies has been developed and is described in the paper.

cs.LG↗

arXiv2023

EvoAAA: An evolutionary methodology for automated \neural autoencoder architecture search

Machine learning models work better when curated features are provided to them. Feature engineering methods have been usually used as a preprocessing step to obtain or build a proper feature set. In late years, autoencoders (a specific type of symmetrical neural network) have been widely used to perform representation learning, proving their competitiveness against classical feature engineering algorithms. The main obstacle in the use of autoencoders is finding a good architecture, a process that most experts confront manually. An automated autoencoder architecture search procedure, based on evolutionary methods, is proposed in this paper. The methodology is tested against nine heterogeneous data sets. The obtained results show the ability of this approach to find better architectures, able to concentrate most of the useful information in a minimized coding, in a reduced time.

cs.NE↗

arXiv2020

New approximate-analytical solutions for the nonlinear fractional Schrödinger equation with second-order spatio-temporal dispersion via double Laplace transform method

In this paper, a modified nonlinear Schrödinger equation with spatio-temporal dispersion is formulated in the senses of Caputo fractional derivative and conformable derivative. A new generalized double Laplace transform coupled with Adomian decomposition method has been defined and applied to solve the newly formulated nonlinear Schrödinger equation with spatio-temporal dispersion. The approximate analytical solutions using the proposed generalized method in the sense of Caputo fractional derivative and conformable derivatives are obtained and compared with each other graphically.

math.GM↗

arXiv2015

Detecting hidden periodicities for models with cyclical errors

In this paper, the estimation of parameters in the harmonic regression with cyclically dependent errors is addressed. Asymptotic properties of the least-squares estimates are analyzed by simulation experiments. By numerical simulation, we prove that consistency and asymptotic normality of the least-squares parameter estimator studied holds under different scenarios, where theoretical results do not exist, and have yet to be proven. In particular, these two asymptotic properties are shown by simulations for the least-squares parameter estimator in the non-linear regression model analyzed, when its error term is defined as a non-linear transformation of a Gaussian random process displaying long-range dependence.

stat.ME↗