Symbolic regression for scientific discovery: an application to wind speed forecasting

Ismail Alaoui Abdellaoui; Siamak Mehrkanoon

doi:10.1109/SSCI50451.2021.9659860

Symbolic regression for scientific discovery: an application to wind speed forecasting

Ismail Alaoui Abdellaoui, Siamak Mehrkanoon^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

Abstract

Symbolic regression corresponds to an ensemble of techniques that allow to uncover an analytical equation from data. Through a closed form formula, these techniques provide great advantages such as potential scientific discovery of new laws, as well as explainability, feature engineering as well as fast inference. Similarly, deep learning based techniques has shown an extraordinary ability of modeling complex patterns. The present paper aims at applying a recent end-to-end symbolic regression technique, i.e. the equation learner (EQL), to get an analytical equation for wind speed forecasting. We show that it is possible to derive an analytical equation that can achieve reasonable accuracy for short term horizons predictions only using few number of features.

Original language	English
Title of host publication	2021 IEEE Symposium Series on Computational Intelligence (SSCI)
Publisher	IEEE
Pages	01-08
Number of pages	8
DOIs	https://doi.org/10.1109/SSCI50451.2021.9659860
Publication status	Published - 2021
Event	2021 IEEE Symposium Series on Computational Intelligence - Online, IEEE, Orlando, United States Duration: 5 Dec 2021 → 7 Dec 2021 https://attend.ieee.org/ssci-2021/

Symposium

Symposium	2021 IEEE Symposium Series on Computational Intelligence
Abbreviated title	IEEE SSCI 2021
Country/Territory	United States
City	Orlando
Period	5/12/21 → 7/12/21
Internet address	https://attend.ieee.org/ssci-2021/

Access to Document

10.1109/SSCI50451.2021.9659860

https://ieeexplore.ieee.org/document/9659860

Cite this

@inproceedings{095c7fd241144e0f83ed5b3730c83369,

title = "Symbolic regression for scientific discovery: an application to wind speed forecasting",

abstract = "Symbolic regression corresponds to an ensemble of techniques that allow to uncover an analytical equation from data. Through a closed form formula, these techniques provide great advantages such as potential scientific discovery of new laws, as well as explainability, feature engineering as well as fast inference. Similarly, deep learning based techniques has shown an extraordinary ability of modeling complex patterns. The present paper aims at applying a recent end-to-end symbolic regression technique, i.e. the equation learner (EQL), to get an analytical equation for wind speed forecasting. We show that it is possible to derive an analytical equation that can achieve reasonable accuracy for short term horizons predictions only using few number of features.",

author = "Abdellaoui, {Ismail Alaoui} and Siamak Mehrkanoon",

note = "Funding Information: All authors have completed and submitted the International Committee of Medical Journal Editors form for disclosure of potential conflicts of interest. John Zervos and Tyler Prentiss report grants from the United Way of Southeastern Michigan, Vattikuti Foundation, and Abbott Laboratories during the conduct of the study. Marcus J. Zervos reports grants from Pfizer, Merck, and Serono, outside the submitted work. No other potential conflicts of interest were disclosed. Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 IEEE Symposium Series on Computational Intelligence, IEEE SSCI 2021 ; Conference date: 05-12-2021 Through 07-12-2021",

year = "2021",

doi = "10.1109/SSCI50451.2021.9659860",

language = "English",

pages = "01--08",

booktitle = "2021 IEEE Symposium Series on Computational Intelligence (SSCI)",

publisher = "IEEE",

address = "United States",

url = "https://attend.ieee.org/ssci-2021/",

}

TY - GEN

T1 - Symbolic regression for scientific discovery: an application to wind speed forecasting

AU - Abdellaoui, Ismail Alaoui

AU - Mehrkanoon, Siamak

N1 - Funding Information: All authors have completed and submitted the International Committee of Medical Journal Editors form for disclosure of potential conflicts of interest. John Zervos and Tyler Prentiss report grants from the United Way of Southeastern Michigan, Vattikuti Foundation, and Abbott Laboratories during the conduct of the study. Marcus J. Zervos reports grants from Pfizer, Merck, and Serono, outside the submitted work. No other potential conflicts of interest were disclosed. Publisher Copyright: © 2021 IEEE.

PY - 2021

Y1 - 2021

N2 - Symbolic regression corresponds to an ensemble of techniques that allow to uncover an analytical equation from data. Through a closed form formula, these techniques provide great advantages such as potential scientific discovery of new laws, as well as explainability, feature engineering as well as fast inference. Similarly, deep learning based techniques has shown an extraordinary ability of modeling complex patterns. The present paper aims at applying a recent end-to-end symbolic regression technique, i.e. the equation learner (EQL), to get an analytical equation for wind speed forecasting. We show that it is possible to derive an analytical equation that can achieve reasonable accuracy for short term horizons predictions only using few number of features.

AB - Symbolic regression corresponds to an ensemble of techniques that allow to uncover an analytical equation from data. Through a closed form formula, these techniques provide great advantages such as potential scientific discovery of new laws, as well as explainability, feature engineering as well as fast inference. Similarly, deep learning based techniques has shown an extraordinary ability of modeling complex patterns. The present paper aims at applying a recent end-to-end symbolic regression technique, i.e. the equation learner (EQL), to get an analytical equation for wind speed forecasting. We show that it is possible to derive an analytical equation that can achieve reasonable accuracy for short term horizons predictions only using few number of features.

U2 - 10.1109/SSCI50451.2021.9659860

DO - 10.1109/SSCI50451.2021.9659860

M3 - Conference article in proceeding

SP - 1

EP - 8

BT - 2021 IEEE Symposium Series on Computational Intelligence (SSCI)

PB - IEEE

T2 - 2021 IEEE Symposium Series on Computational Intelligence

Y2 - 5 December 2021 through 7 December 2021

ER -