Classification in conservation biology: A comparison of five machine-learning methods

Kampichler, Christian; Wieland, Ralf; Calmé, Sophie; Weissenberger, Holger; Arriaga-Weiss, Stefan L.

doi:10.1016/j.ecoinf.2010.06.003

Cited by 162 publications

(123 citation statements)

References 45 publications

Supporting

Mentioning

121

Contrasting

Unclassified

Order By: Relevance

“…They found that the tree-based models were generally superior to predict species richness of reef fish. Furthermore, Mouton et al (2011) found similar predictive performance of RF and Fuzzy logic models to represent mesohabitat suitability for Salmo trutta in Spain, whereas Kampichler et al (2010) compared different ML techniques (including ANN and RF) for classification problems and recommend the use of RF in conservation biology. Given the large number of ML techniques, there are not established guidelines for defining the most appropriate method to address a particular ecological question or management action for freshwater ecosystems.…”

Section: Introductionmentioning

confidence: 94%

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

Olaya-Marín

Martínez‐Capel

Vezza

2013

Knowl. Managt. Aquatic Ecosyst.

View full text Add to dashboard Cite

Key-words:Artificial neural networks, random forests, native fish, species richness, Mediterranean rivers Machine learning (ML) techniques have become important to support decision making in management and conservation of freshwater aquatic ecosystems. Given the large number of ML techniques and to improve the understanding of ML utility in ecology, it is necessary to perform comparative studies of these techniques as a preparatory analysis for future model applications. The objectives of this study were (i) to compare the reliability and ecological relevance of two predictive models for fish richness, based on the techniques of artificial neural networks (ANN) and random forests (RF) and (ii) to evaluate the conformity in terms of selected important variables between the two modelling approaches. The effectiveness of the models were evaluated using three performance metrics: the determination coefficient (R 2 ), the mean squared error (MSE) and the adjusted determination coefficient (R 2 adj ) and both models were developed using a k-fold crossvalidation procedure. According to the results, both techniques had similar validation performance (R 2 = 68% for RF and R 2 = 66% for ANN). Although the two methods selected different subsets of input variables, both models demonstrated high ecological relevance for the conservation of native fish in the Mediterranean region. Moreover, this work shows how the use of different modelling methods can assist the critical analysis of predictions at a catchment scale. RÉSUMÉUne comparaison des réseaux de neurones et des forêts aléatoires pour prédire la richesse en espèces de poissons indigènes dans les rivières méditerranéennes Les techniques d'apprentissage automatique (ML) sont devenues importantes pour aider à la décision dans la gestion et la conservation des écosystèmes aquatiques d'eau douce. Étant donné le grand nombre de techniques ML pour amélio-rer la compréhension de l'utilité des ML en écologie, il est nécessaire de réaliser des études comparatives de ces techniques comme analyse préparatoire pour des applications de modèles futurs. Les objectifs de cette étude étaient : (i) de comparer la fiabilité et la pertinence écologique de deux modèles prédictifs pour la richesse de poisson, basé sur les techniques de réseaux de neurones artificiels (ANN) et les forêts aléatoires (RF) et (ii) d'évaluer la conformité en termes de sélection des variables importantes entre les deux approches de modélisa-tion. L'efficacité des modèles a été évaluée au moyen de trois indicateurs de Article published by EDP Sciences E.J. Olaya-Marín et al.: Knowl. Managt. Aquatic Ecosyst. (2013) 409, 07 performance : le coefficient de détermination (R 2 ), l'erreur quadratique moyenne (MSE) et le coefficient de détermination ajusté (R 2 adj ) et les deux modèles ont été développés en utilisant une procédure de validation croisée k-fold. Selon les résul-tats, les deux techniques ont des performances de validation similaires (R 2 = 68 % pour RF et R 2 = 66 % pour ANN). Bien que les deux...

show abstract

Section: Introductionmentioning

confidence: 94%

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

Olaya-Marín

Martínez‐Capel

Vezza

2013

Knowl. Managt. Aquatic Ecosyst.

View full text Add to dashboard Cite

show abstract

“…Similar comparative studies that attempted to examine accuracy differences among different machine learning algorithms showed different results. Studies by Adam, Mutanga [46] and Duro, Franklin [22] indicated machine learning algorithms such as SVM and RF performed similarly, and other studies such as that of Cracknell and Reading [47], Fernández-Delgado, Cernadas [17] and Kampichler, Wieland [48] showed RF to be the most accurate algorithm. Our result, which showed SVM to be superior in accuracy compared to both RF and C5.0, is yet in agreement with studies such as that of Shao and Lunetta [16] and Maroco, Silva [49] indicated SVM to be superior in accuracy.…”

Section: R E T R a C T E Dmentioning

confidence: 99%

RETRACTED: <i>A Participatory Iterative Mapping Approach and Evaluation of Three Machine Learning Algorithms for Accurate Mapping of Cropping Patterns in a Complex Agro-Ecosystems</i>

Feyisa¹,

Palao²,

Nelson³

et al. 2016

ARS

View full text Add to dashboard Cite

“…It's high predictive power has been supported by previous comparative studies with other machine learning (ML) methods [19][20][21]. The final classification is obtained by combining the classification results from the individual decision trees.…”

Section: Selection Of the Pathways Pairsmentioning

confidence: 94%

Revealing pathway cross-talk related to diabetes mellitus by Monte Carlo Cross-Validation analysis

Cai

Shi

2017

Open Life Sciences

View full text Add to dashboard Cite

Abstract:Objective: To explore potential functional biomarkers in diabetes mellitus (DM) by utilizing gene pathway cross-talk. Methods: Firstly, potential disrupted pathways that were enriched by differentially expressed genes (DEGs) were identified based on biological pathways downloaded from the Ingenuity Pathways Analysis (IPA) database. In addition, we quantified the pathway crosstalk for each pair of pathways based on Discriminating Score (DS). Random forest (RF) classification was then employed to find the top 10 pairs of pathways with a high area under the curve (AUC) value between DM samples versus normal samples based on 10-fold cross-validation. Finally, a Monte Carlo Cross-Validation was applied to demonstrate the identified pairs of pathways by a mutual information analysis. Results: A total of 247 DEGs in normal and disease samples were identified. Based on the F-test, 50 disrupted pathways were obtained with false discovery rate (FDR) < 0.01. Simultaneously, after calculating the DS, the top 10 pairs of pathways were selected based on a higher AUC value as measured by RF classification. From the Monte Carlo Cross-Validation, we considered the top 10 pairs of pathways with higher AUC values ranked for all 50 bootstraps as the most frequently detected ones. Conclusion: The pairs of pathways identified in our study might be key regulators in DM.

show abstract

Classification in conservation biology: A comparison of five machine-learning methods

Cited by 162 publications

References 45 publications

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

RETRACTED: <i>A Participatory Iterative Mapping Approach and Evaluation of Three Machine Learning Algorithms for Accurate Mapping of Cropping Patterns in a Complex Agro-Ecosystems</i>

Revealing pathway cross-talk related to diabetes mellitus by Monte Carlo Cross-Validation analysis

Contact Info

Product

Resources

About

Classification in conservation biology: A comparison of five machine-learning methods

Cited by 162 publications

References 45 publications

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

A comparison of artificial neural networks and random forests to predict native fish species richness in Mediterranean rivers

RETRACTED: &lt;i&gt;A Participatory Iterative Mapping Approach and Evaluation of Three Machine Learning Algorithms for Accurate Mapping of Cropping Patterns in a Complex Agro-Ecosystems&lt;/i&gt;

Revealing pathway cross-talk related to diabetes mellitus by Monte Carlo Cross-Validation analysis

Contact Info

Product

Resources

About

RETRACTED: <i>A Participatory Iterative Mapping Approach and Evaluation of Three Machine Learning Algorithms for Accurate Mapping of Cropping Patterns in a Complex Agro-Ecosystems</i>