Next Article in Journal
Is the Synthetic Fungicide Fosetyl-Al Safe for the Ecotoxicological Models Danio rerio and Enchytraeus crypticus?
Previous Article in Journal
Support Group as a Management Modality for Burning Mouth Syndrome: A Randomized Prospective Study
 
 
Article

Evaluation of Machine Learning Predictions of a Highly Resolved Time Series of Chlorophyll-a Concentration

1
Helmholtz Centre for Polar and Marine Research, Alfred Wegener Institute, Wadden Sea Research Station, Hafenstr. 43, 25992 List auf Sylt, Germany
2
Division of Climate Sciences, Section of Paleoclimate Dynamics, Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research, 27570 Bremerhaven, Germany
3
Department of Environmental Physics & MARUM, University of Bremen, 28359 Bremen, Germany
*
Author to whom correspondence should be addressed.
Academic Editor: Fabio La Foresta
Appl. Sci. 2021, 11(16), 7208; https://doi.org/10.3390/app11167208
Received: 16 July 2021 / Revised: 30 July 2021 / Accepted: 2 August 2021 / Published: 5 August 2021
(This article belongs to the Section Ecology Science and Engineering)
Pelagic chlorophyll-a concentrations are key for evaluation of the environmental status and productivity of marine systems, and data can be provided by in situ measurements, remote sensing and modelling. However, modelling chlorophyll-a is not trivial due to its nonlinear dynamics and complexity. In this study, chlorophyll-a concentrations for the Helgoland Roads time series were modeled using a number of measured water and environmental parameters. We chose three common machine learning algorithms from the literature: the support vector machine regressor, neural networks multi-layer perceptron regressor and random forest regressor. Results showed that the support vector machine regressor slightly outperformed other models. The evaluation with a test dataset and verification with an independent validation dataset for chlorophyll-a concentrations showed a good generalization capacity, evaluated by the root mean squared errors of less than 1 µg L−1. Feature selection and engineering are important and improved the models significantly, as measured in performance, improving the adjusted R2 by a minimum of 48%. We tested SARIMA in comparison and found that the univariate nature of SARIMA does not allow for better results than the machine learning models. Additionally, the computer processing time needed was much higher (prohibitive) for SARIMA. View Full-Text
Keywords: time series regression; artificial intelligence; Helgoland Roads time series; support vector machine; multi-layer perceptron; random forest; productivity; SARIMA time series regression; artificial intelligence; Helgoland Roads time series; support vector machine; multi-layer perceptron; random forest; productivity; SARIMA
Show Figures

Figure 1

MDPI and ACS Style

Amorim, F.d.L.L.d.; Rick, J.; Lohmann, G.; Wiltshire, K.H. Evaluation of Machine Learning Predictions of a Highly Resolved Time Series of Chlorophyll-a Concentration. Appl. Sci. 2021, 11, 7208. https://doi.org/10.3390/app11167208

AMA Style

Amorim FdLLd, Rick J, Lohmann G, Wiltshire KH. Evaluation of Machine Learning Predictions of a Highly Resolved Time Series of Chlorophyll-a Concentration. Applied Sciences. 2021; 11(16):7208. https://doi.org/10.3390/app11167208

Chicago/Turabian Style

Amorim, Felipe de Luca Lopes de, Johannes Rick, Gerrit Lohmann, and Karen Helen Wiltshire. 2021. "Evaluation of Machine Learning Predictions of a Highly Resolved Time Series of Chlorophyll-a Concentration" Applied Sciences 11, no. 16: 7208. https://doi.org/10.3390/app11167208

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop