A hybrid of Random Forest and Deep Auto-Encoder with support vector regression methods for accuracy improvement and uncertainty reduction of long-term streamflow prediction

Abbasi, Mahdi; Farokhnia, Ashkan; Bahreinimotlagh, Masoud; Roozbahani, Reza

doi:10.1016/j.jhydrol.2020.125717

Cited by 62 publications

(17 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RFs are appealing for a multitude of reasons, including their predictive performance, robustness, speed, nonparametric nature, stability, diagnosis of variable importance, as well as their ability to handle nonlinearity, interactions, noise, and small sample sizes in forcing data (Tyralis et al., 2019). RFs are being used in a variety of applications including water resources and water quality modeling (Suchetana et al., 2017), construction safety risk (Tixier et al., 2016), and used with success in recent years for flow forecasting, primarily at daily and monthly time scales (Abbasi et al., 2021; Al‐Juboori, 2019; Ghorbani et al., 2020; Hussain & Khan, 2020; Li et al., 2019; Liang et al., 2018; Muñoz et al., 2018; Papacharalampous & Tyralis, 2018; Pham et al., 2020).…”

Section: Proposed Methodologymentioning

confidence: 99%

Stochastic Decadal Projections of Colorado River Streamflow and Reservoir Pool Elevations Conditioned on Temperature Projections

Woodson

Rajagopalan

Baker

et al. 2021

Water Resources Research

View full text Add to dashboard Cite

Decadal (∼10‐year)‐scale flow projections in the Colorado River Basin (CRB) are increasingly important for water resources management and planning of its reservoir system. Physical models, ensemble streamflow prediction (ESP), do not have skill beyond interannual time scales. However, Global Climate Models have good skill in projecting decadal temperatures. This, combined with the sensitivity of CRB flows to temperature from recent studies, motivate the research question, can skill in decadal temperature projections be translated to operationally skillful flow projections and consequently, water resources management? To explore this, we used temperature projections from the Community Earth System Model‐Decadal Prediction Large Ensemble (CESM‐DPLE) along with past basin runoff efficiency as covariates in a Random Forest (RF) method to project ensembles of multiyear mean flow at the key aggregate gauge of Lees Ferry, Arizona. RF streamflow projections outperformed both ESP and climatology in a 1982–2017 hindcast, as measured by ranked probability skill score. The projections were disaggregated to monthly and subbasin scales to drive the Colorado River Mid‐term Modeling System (CRMMS) to generate ensembles of water management variables. The projections of pool elevations in Lakes Powell and Mead, the two largest U.S. reservoirs that are critical for water resources management in the basin, were found to reduce the hindcast median root mean square error by up to −20% and −30% at lead times of 48 and 60 months, respectively, relative to projections generated from ESP. This suggests opportunities for enhancing water resources management in the CRB and potentially elsewhere.

show abstract

Section: Proposed Methodologymentioning

confidence: 99%

Stochastic Decadal Projections of Colorado River Streamflow and Reservoir Pool Elevations Conditioned on Temperature Projections

Woodson

Rajagopalan

Baker

et al. 2021

Water Resources Research

View full text Add to dashboard Cite

show abstract

“…The Autoencoder (AE) [ 21 ] and Sparse Autoencoder (SAE) networks [ 22 ] are such deep learning algorithms. Because of this, some researchers [ 23 , 24 ] adopted AE networks to compress high-dimensional hydrological time series to obtain low-dimensional representations and generate predictions based on low-dimensional representations. Besides, others researchers [ 25 – 27 ] exploited the SAE network to extract features from hydrological time series and generate predictions based on the extracted features.…”

Section: Introductionmentioning

confidence: 99%

Prediction model of sparse autoencoder-based bidirectional LSTM for wastewater flow rate

Huang

Yang

et al. 2022

J Supercomput

View full text Add to dashboard Cite

“…There are numerous data-driven models, which have been used for time series forecasting like Multiple Linear Regression (MLR) (Abbasi et al 2021), K-Nearest Neighbor (KNN) (Modaresi et al 2018), and Arti cial Neural Networks (ANN) (Khazaee Poul et al 2019). The mentioned models are the classic Data-Driven Forecasting Frameworks (DDFF).…”

Section: Introductionmentioning

confidence: 99%

A New Criteria for Determining the Best Decomposition Level and Filter in Wavelet-based Data-driven Forecasting Frameworks- Validating Using Three Case studies on The CAMELS Dataset

Behbahani

2022

Preprint

View full text Add to dashboard Cite

Recently, numerous papers have been published in the field of using preprocessing models (e.g. Discrete wavelet) in Data-driven Forecasting Frameworks (DDFF). There are some unresolved problems in these models like using future data, boundary affected data, and miss selection of decomposition level and wavelet filter that cause an erroneous result. However, Wavelet-based Data-driven Forecasting Framework (WDDFF) solves these problems. The first two problems could be solved using Maximal Overlap Discrete Wavelet Transform (MODWT) and a trous algorithm (AT). As the best we know, there is no absolute solution for decomposition level and wavelet filter selection. Meanwhile, as a novel investigation, we are going to use Entropy to find a solution for these problems. We are using the concept of predictability of time series using entropy for determining decomposition level and suitable filter, to develop the Maximal Overlap Discrete Wavelet-Entropy Transform (MODWET) to apply in WDDFF correctly. We will reveal the effectiveness of MODWET through three real-world case studies on the CAMELS data set. In these case studies, we will forecast the streamflow of determined stations from one month ahead to prove the effectiveness of using preprocessing models on forecasting accuracy. The proposed model is a combination of Input Variable Selection (IVS), preprocessing model, and Data-Driven Model (DDM). In conclusion, we will show that MODWET- ANN is the best model. In addition, we will realize how good entropy could find decomposition level and filter, which solves the mentioned concerns about using WDDFF in real-world hydrological forecasting problems.

show abstract

A hybrid of Random Forest and Deep Auto-Encoder with support vector regression methods for accuracy improvement and uncertainty reduction of long-term streamflow prediction

Cited by 62 publications

References 36 publications

Stochastic Decadal Projections of Colorado River Streamflow and Reservoir Pool Elevations Conditioned on Temperature Projections

Stochastic Decadal Projections of Colorado River Streamflow and Reservoir Pool Elevations Conditioned on Temperature Projections

Prediction model of sparse autoencoder-based bidirectional LSTM for wastewater flow rate

A New Criteria for Determining the Best Decomposition Level and Filter in Wavelet-based Data-driven Forecasting Frameworks- Validating Using Three Case studies on The CAMELS Dataset

Contact Info

Product

Resources

About