Long short-term memory (LSTM) recurrent neural network for low-flow hydrological time series forecasting

Sahoo, Bibhuti Bhusan; Jha, Ramakar; Singh, Anshuman; Kumar, Deepak

doi:10.1007/s11600-019-00330-1

Cited by 228 publications

(88 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Studies have demonstrated a wide range of forecasting accuracies using AI models with the geographic topographies of the test sites despite their success in other fields [15][16][17][18]. e development of a model that will be used to forecast in any environmental condition is an aspect yet to be explored.…”

Section: Introductionmentioning

confidence: 99%

Hybridized Extreme Learning Machine Model with Salp Swarm Algorithm: A Novel Predictive Model for Hydrological Application

Yaseen‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

Faris

Al‐Ansari

2020

Complexity

View full text Add to dashboard Cite

e capability of the extreme learning machine (ELM) model in modeling stochastic, nonlinear, and complex hydrological engineering problems has been proven remarkably. e classical ELM training algorithm is based on a nontuned and random procedure that might not be efficient in convergence of excellent performance or possible entrapment in the local minima problem. is current study investigates the integration of a newly explored metaheuristic algorithm (i.e., Salp Swarm Algorithm (SSA)) with the ELM model to forecast monthly river flow. Twenty years of river flow data time series of the Tigris river at the Baghdad station, Iraq, is used as a case study. Different input combinations are applied for constructing the predictive models based on antecedent values. e results are evaluated based on several statistical measures and graphical presentations. e river flow forecast accuracy of SSA-ELM outperformed the classical ELM and other artificial intelligence (AI) models. Over the testing phase, the proposed SSA-ELM model yielded a satisfactory enhancement in the level accuracies (8.4 and 13.1 percentage of augmentation for RMSE and MAE, respectively) against the classical ELM model. In summary, the study ascertains that the SSA-ELM model is a qualified data-intelligent model for monthly river flow prediction at the Tigris river, Iraq.

show abstract

Section: Introductionmentioning

confidence: 99%

Hybridized Extreme Learning Machine Model with Salp Swarm Algorithm: A Novel Predictive Model for Hydrological Application

Yaseen‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬‬

Faris

Al‐Ansari

2020

Complexity

View full text Add to dashboard Cite

show abstract

“…the covariances between the pairwise temporal independent variables are assumed to be equal, based on the distance (in days) between the lags 8 . 8 Hence, for the covariances between successive days:…”

Section: Hypothesesmentioning

confidence: 99%

“…For example, if a temporary process was measured in days, and the maximum delay to accurately forecast the series was five lags, it would follow that the memory of the series, this is, the autoregressive process (AR), is 5 days (AR (5)). This statistical concept is increasingly used in time series as applied in physical, social, engineering, and statistical sciences (5)(6)(7)(8). Indeed, the designers of time series analysis have used this notion to define short and long memory processes (9).…”

Section: Introductionmentioning

confidence: 99%

Pooled Time Series Modeling Reveals Smoking Habit Memory Pattern

et al. 2020

View full text Add to dashboard Cite

Smoking is a habit that is hard to break because nicotine is highly addictive and smoking behavior is strongly linked to multiple daily activities and routines. Here, we explored the effect of gender, age, day of the week, and previous smoking on the number of cigarettes smoked on any given day. Data consisted of daily records of the number of cigarettes participants smoked over an average period of 84 days. The sample included smokers (36 men and 26 women), aged between 18 and 26 years, who smoked at least five cigarettes a day and had smoked for at least 2 years. A panel data analysis was performed by way of multilevel pooled time series modeling. Smoking on any given day was a function of the number of cigarettes smoked on the previous day, and 2, 7, 14, 21, 28, 35, 42, 49, and 56 days previously, and the day of the week. Neither gender nor age influenced this pattern, with no multilevel effects being detected, thus the behavior of all participants fitted the same smoking model. These novel findings show empirically that smoking behavior is governed by firmly established temporal dependence patterns and inform temporal parameters for the rational design of smoking cessation programs.

show abstract

“…Bastos 2010 Com isso, fica claro que a previsão de tributos utilizando modelos baseados em aprendizado profundo, ainda não devidamente explorado e carecendo de maiores contribuições, principalmente a partir dos novos modelos baseados em redes recorrentes, que têm tido bastante sucesso em tantas outras áreas de aplicação [Schmidhuber 2015], [Namin and Namin 2018], [Sahoo et al 2019] e [Khodabakhsh et al 2020].…”

Section: Trabalhos Relacionadosunclassified

Aprendizado Profundo Aplicado na Previsão de Receita Tributária Utilizando Variáveis Endógenas

Silva¹,

Figueiredo²

2020

Anais Do Encontro Nacional De Inteligência Artificial E Computacional (ENIAC 2020)

View full text Add to dashboard Cite

A previsão de receita tributária para a gestão de um estado em situação de crise econômico-financeira, caso do estado do Rio de Janeiro, tornou-se uma tarefa fundamental e desafiadora para as secretarias de fazenda e planejamento dos Estados, pois há incertezas econômicas e políticas que afetam a série temporal. Nesse sentido, este trabalho inicia o projeto de investigação e uso de modelos mais novos e acurados de Machine Learning, como Long Short-Term Memory (LSTM), para prever a receita de tributos. Os resultados apresentam erro relativo menor que 1%, para previsão do ICMS, indicando desempenho superior às previsões realizadas pela SEFAZ-RJ e aos modelos MLP, usados para efeitos de comparação.

show abstract

Long short-term memory (LSTM) recurrent neural network for low-flow hydrological time series forecasting

Cited by 228 publications

References 66 publications

Hybridized Extreme Learning Machine Model with Salp Swarm Algorithm: A Novel Predictive Model for Hydrological Application

Hybridized Extreme Learning Machine Model with Salp Swarm Algorithm: A Novel Predictive Model for Hydrological Application

Pooled Time Series Modeling Reveals Smoking Habit Memory Pattern

Aprendizado Profundo Aplicado na Previsão de Receita Tributária Utilizando Variáveis Endógenas

Contact Info

Product

Resources

About