Smoothing Strategies Combined with ARIMA and Neural Networks to Improve the Forecasting of Traffic Accidents

Barba, Lida; Rodríguez, Nibaldo; Montt, Cecilia

doi:10.1155/2014/152375

Cited by 26 publications

(21 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…El preprocesamiento de los datos por medio de SSA ha sido implementado a partir de una longitud inicial de ventana L=N/2, donde N es el tamaño de la muestra. La energía de los valores propios obtenidos en el segundo paso de SSA se representa de forma gráfica para encontrar el tamaño de ventana efectiva (Barba L. et al 2014), en este caso L=6. El embebido vuelve a realizarse con este valor efectivo, la componente extraída muestra las oscilaciones de baja frecuencia; mientras que la componente xa muestra el comportamiento periódico de alta frecuencia, como se muestra en la Figura 2a.…”

Section: Resultados Y Discusiónunclassified

“…Singular Spectrum Analysis and Autoregressive models... Boletín Geológico y Minero, 129 ( tial window length used was computed with the general window length value L=N/2; where N is the sample size (N=156). The differential energy of the eigen-values obtained in the second step of SSA was plotted to find the highest relative energy concentration (Barba L. et al, 2014); this is shown in Figure 1b. From Figure 1b, the highest peak L=6 was used as the effective window length.…”

Section: Preprocessing With Singular Spectrum Analysismentioning

confidence: 99%

See 1 more Smart Citation

Singular Spectrum Analysis and Autoregressive models for Ecuadorian shrimp catch forecasting

Barba¹,

Rodríguez²

2018

Bol. geol. min.

View full text Add to dashboard Cite

Shrimp trawl fishing is a relevant activity that contributes significantly to the generation of economic resources for Ecuador through the creation of jobs and the commercialization of the shrimps. Unfortunately the high variability of the marine ecosystem signals are complex making forecasting a difficult task. This paper evaluates the use of singular spectrum analysis (SSA) for improving forecasting by conventional methods. SSA is implemented in four steps, embedding, decomposition, grouping, and diagonal averaging. The effective length of the embedding window is selected by means of the information provided by the differential energy of the eigenvalues. The SSA decomposition is used to extract two types of components, one of low frequency which represents the inter-annual component, and the other of high frequency which represents the annual component. Once the components have been obtained, three models based on SSA are implemented, the first is based on the autoregressive model (SSA-AR), the second is an artificial neural network (ANN) based on the Levenberg-Marquardt algorithm (SSA-ANN-LM), and the third is another ANN based on the Cuckoo Search algorithm (SSA-ANN-CS). The historical data that was used to evaluate the models are the shrimp monthly catches in the Gulf of Guayaquil from 2000 to 2012. The empirical results show the superiority of the proposed models with respect to the conventional models. The best approximation was achieved with the enhanced model SSA-AR, with a MAPE of 1.0%, a R 2 of 99% and a RMSE of 1.5%.El aprendizaje de la red es evaluado por medio de dos algoritmos, uno basado en el descenso del gradiente (Levenberg-Marquardt) y otro basado en una metaheurística (Cuckoo Search). Métricas de RendimientoLa exactitud del pronóstico es evaluada por medio de las métricas: Mean Absolute Percentage Error (MAPE),

show abstract

Section: Resultados Y Discusiónunclassified

Section: Preprocessing With Singular Spectrum Analysismentioning

confidence: 99%

Singular Spectrum Analysis and Autoregressive models for Ecuadorian shrimp catch forecasting

Barba¹,

Rodríguez²

2018

Bol. geol. min.

View full text Add to dashboard Cite

show abstract

“…Many studies have shown their performance as separate forecasting models and combined with each other. Interesting results on different nature of time series are presented by Wang L. et al [29]; Tseng et al [37]; Sheta et al [38]; Barba et al [39]. In the field of energy forecasting the studies are focused on choosing one of the models among some of them as has been discussed by many authors [8-9, 11-18, 40-43].…”

Section: Baseline Scenario Forecastingmentioning

confidence: 99%

A Study on the Efficiency of Hybrid Models in Forecasting Precipitations and Water Inflow Albania Case Study

Gjika¹,

Ferrja²,

Kamberi³

2019

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

Climatic changes have a significant impact on many real life processes. Climacteric position of Albania makes precipitations and water inflows in HPP the main variables influencing the amount of electric energy produced in the country. Taking into account their volatility it has considerably increased the need of using hybrid models to improve the quality of predictions. After a detailed analysis of the time series components, we develop a group of hybrid models and propose modifications to increase the accuracy in prediction. Among the contributions of this work is the challenge to choose between hybrid models presented earlier in literature and the modified version according to the nature of data. The final decision on the most accurate model is made based on many goodness of fit test. This study suggest that an accurate selection of the forecasting techniques increases significantly the quality of forecast on precipitations and water inflow data.

show abstract

“…Barba et al combined ARIMA method and artificial neural network to improve the accuracy of futures studies model. Their results show that combining the two methods provides values with high accuracy of prediction [10]. Prediction is usually based on time series, but collecting data in a timely manner takes a lot of time and money [11].…”

Section: Introductionmentioning

confidence: 99%

Estimating Accident-Related Traumatic Injury Rate by Future Studies Models in Semnan Province, Iran

Omidi

2018

Health Emerg. Disast.

View full text Add to dashboard Cite

Background: Any accident is a disturbance in the balance between the human system, vehicle, road and environment. Future prediction of traumatic accidents is a valuable factor for managers to make strategic decisions in the areas of safety, health and transportation. Materials and Methods: In this study, by using Grey Model (GM) (1.1), Rolling Grey Model (RGM), Fourier Grey Model (FGM) (1.1), survival modification model, ARIMA time series, harmonic pattern and statistical data, the number of traffic injuries referred to forensic medicine centers in Semnan Province between 2017 and 2020 were predicted based on the number of traffic injured in Semnan Province from March 2009 and March 2016. Results: The mean absolute error percentage for the GM (1.1), RGM (1), FGM (1.1), survival model, ARIMA and harmonic models were 0.994, 0.082, 0.091, 0.105, 0.05, 0.11, respectively, indicating a greater accuracy of the ARIMA method, compared to the other methods. The number of road traffic injuries in Semnan Province is decreasing and will reach 4052 in 2020. Conclusion: ARIMA model is the best method of the future studies model for the number of injured patients referred to the forensic medicine centers in Semnan Province compared to other studied methods. Future studies model shows that the injuries caused by accidents in the province of Semnan are decreasing

show abstract

Smoothing Strategies Combined with ARIMA and Neural Networks to Improve the Forecasting of Traffic Accidents

Cited by 26 publications

References 40 publications

Singular Spectrum Analysis and Autoregressive models for Ecuadorian shrimp catch forecasting

Singular Spectrum Analysis and Autoregressive models for Ecuadorian shrimp catch forecasting

A Study on the Efficiency of Hybrid Models in Forecasting Precipitations and Water Inflow Albania Case Study

Estimating Accident-Related Traumatic Injury Rate by Future Studies Models in Semnan Province, Iran

Contact Info

Product

Resources

About