Simplifying artificial neural network models of river basin behaviour by an automated procedure for input variable selection

Oliveira, Guilherme Garcia de; Pedrollo, Olavo Correa; Castro, Nilza Maria dos Reis

doi:10.1016/j.engappai.2015.01.001

Cited by 27 publications

(15 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…These results are in agreement with the findings of Silva Júnior, Tucci, Castro, and Goldenfun () that small amounts of precipitation do not produce local surface runoff and often lead to increased subsurface flows in this catchment because of the gently sloped terrain. In the last two decades, regional frequency analysis and rainfall‐runoff modeling approaches have been used to investigate changes in hydrological processes at scales ranging from 0.1 to 10 4 km 2 (Girardi, Castro, Goldenfun, Silveira, & Pinheiro, ; Oliveira, Pedrollo, & Castro, ; Silva Júnior et al, ) and to develop water availability models for ungauged catchments.…”

Section: Methodsmentioning

confidence: 99%

Fluvial carbon export and CO₂ efflux in representative nested headwater catchments of the eastern La Plata River Basin

Sorribas

Marques

Castro

et al. 2017

Hydrological Processes

View full text Add to dashboard Cite

This study involved a baseline evaluation of fluvial carbon export and degas rates in three nested rural catchments (1 to 80 km2) in Taboão, a representative experimental catchment of the Upper Uruguay River Basin. Analyses of the carbon content in stream waters and the catchment carbon yield were based on 4‐year monthly in situ data and statistical modeling using the United States Geological Survey load estimator model. We also estimated pCO2 and degas fluxes using carbonate equilibrium and gas‐exchange formulas. Our results indicated that the water was consistently pCO2 saturated (~90% of the cases) and that the steep terrain favors high gas evasion rates. The mean calculated fluvial export was 5.4 tC·km−2·year−1 with inorganic carbon dominating (dissolved inorganic carbon:dissolved organic carbon ratio >4), and degas rates (~40 tC km−2·year−1) were nearly sevenfold higher than the downstream export. The homogeneous land use in this nested catchment system results in similar water‐quality characteristics, and therefore, export rates are expected to be closely related to the rainfall–runoff relationships at each scale. Although the sampling campaigns did not fully reproduce storm‐event conditions and related effects such as flushing or dilution of in‐stream carbon, our results indicated a potential link between dissolved inorganic carbon and slower hydrological pathways related to subsurface water storage and movement.

show abstract

Section: Methodsmentioning

confidence: 99%

Fluvial carbon export and CO₂ efflux in representative nested headwater catchments of the eastern La Plata River Basin

Sorribas

Marques

Castro

et al. 2017

Hydrological Processes

View full text Add to dashboard Cite

show abstract

“…To simplify the interpretation of multi-layered neural network, we focus on relations between inputs and outputs by treating collectively all intermediate layers. This is because in many applications, we must examine how input variables (neurons) are related to the corresponding outputs [15], [16], [17]. Thus, we try to estimate how input neurons have influences on outputs by considering all intermediate layers.…”

Section: A Problem Of Collective Interpretationmentioning

confidence: 99%

Direct Potentiality Assimilation for Improving Multi-Layered Neural Networks

Kamimura¹

2017

Annals of Computer Science and Information Systems

View full text Add to dashboard Cite

Abstract-The present paper aims to propose a new potential learning method to overcome the problem of collective interpretation for interpreting multi-layered neural networks. The potential learning has been introduced to detect important components of neural networks and to train them, taking into account the importance of components. Recently, it has been applied to multilayered neural networks and then the interpretation of input neurons or variables can be possible by collectively treating intermediate layers. However, the collective interpretation for multi-layered neural networks tends to be instable, because the potentialities computed in the pre-training become different from those in the main training. To overcome this problem, we introduce the potential learning with direct potential assimilation. The direct potential assimilation means that the potentiality assimilation is not applied in the phase of pre-training but it is applied directly to training multi-layered neural networks. The new method was applied to the student evaluation data set. Then, it was observed that the selectivity of connection weights could be increased. Then, the input-output potentiality was quite similar to the regression coefficients of logistic regression analysis. Finally, the new method could extract more explicitly inputoutput relations than the regression coefficients by the logistic regression analysis, while improving generalization performance.

show abstract

“…The ANN hydrological model used was performed in the study of Oliveira et al (2014), and resulted in the application of an algorithm for simplification of the neural network (Oliveira et al, 2015b). The reduction of input variables and neurons in the internal layer was performed using an algorithm that looks at the model performance after the imposition of small disturbances in the ANN input data.…”

Section: Hydrological Simulation Using Artificial Neural Network (Anns)mentioning

confidence: 99%

Stochastic approach to analyzing the uncertainties and possible changes in the availability of water in the future based on scenarios of climate change

Oliveira

Pedrollo

Castro

2015

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. The objective of this study was to analyze the changes and uncertainties related to water availability in the future (for the purposes of this study, the period between 2011 and 2040 was adopted), using a stochastic approach, taking as reference a climate projection from climate model Eta CPTEC/HadCM3. The study was applied to the Ijuí River basin in the south of Brazil. The set of methods adopted involved, among others, correcting the climatic variables projected for the future, hydrological simulation using artificial neural networks (ANNs) to define a number of monthly flows and stochastic modeling to generate 1000 hydrological series with equal probability of occurrence. A multiplicative type stochastic model was developed in which monthly flow is the result of the product of four components: (i) long-term trend component; (ii) cyclic or seasonal component; (iii) time-dependency component; and (iv) random component. In general, the results showed a trend to increased flows. The mean flow for a long period, for instance, presented an alteration from 141.6 m 3 s −1

show abstract

Simplifying artificial neural network models of river basin behaviour by an automated procedure for input variable selection

Cited by 27 publications

References 41 publications

Fluvial carbon export and CO₂ efflux in representative nested headwater catchments of the eastern La Plata River Basin

Fluvial carbon export and CO₂ efflux in representative nested headwater catchments of the eastern La Plata River Basin

Direct Potentiality Assimilation for Improving Multi-Layered Neural Networks

Stochastic approach to analyzing the uncertainties and possible changes in the availability of water in the future based on scenarios of climate change

Contact Info

Product

Resources

About

Simplifying artificial neural network models of river basin behaviour by an automated procedure for input variable selection

Cited by 27 publications

References 41 publications

Fluvial carbon export and CO2 efflux in representative nested headwater catchments of the eastern La Plata River Basin

Fluvial carbon export and CO2 efflux in representative nested headwater catchments of the eastern La Plata River Basin

Direct Potentiality Assimilation for Improving Multi-Layered Neural Networks

Stochastic approach to analyzing the uncertainties and possible changes in the availability of water in the future based on scenarios of climate change

Contact Info

Product

Resources

About

Fluvial carbon export and CO₂ efflux in representative nested headwater catchments of the eastern La Plata River Basin

Fluvial carbon export and CO₂ efflux in representative nested headwater catchments of the eastern La Plata River Basin