Comparison of Support Vector Machines (SVM) and Autoregressive integrated moving average (ARIMA) in daily flow forecasting
Journal Title: Journal of River Engineering - Year 2013, Vol 1, Issue 1
Abstract
In recent years, Computational Intelligence (CI) is applied to solve problems for some physical processes with nonlinear relations. Use of data, extraction of relations between them and generalizing in other situations are the basic of intelligent method. Most important methods such as: artificial neural network, fuzzy logic, genetic algorithm and a newer one, called support vector machine (SVM) are used. Support vector machine (SVM) is one of the new methods that has attracted many researchers in various scientific fields.(ARIMA) model has found a widespread application in many practical sciences. An Autoregressive integrated moving average (ARIMA) model is a generalization of an Autoregressive Moving Average (ARMA) model.In this study the ability of Autoregressive Integrated Moving Average (ARIMA) models in daily river flows in north of Iran is estimated. This paper compares two expert models in daily flow forecasting. The support vector machine (SVM) and Autoregressive integrated moving average (ARIMA) are used to forecast daily river flows in north of Iran and the results of these models are compared with the observed daily values. The observed data that are used in this study stared from 1992 to 2010 for18 years period (6550 days). By comparing root mean square error (RMSE) of the model, it was determined that (SVM) model can forecast daily river flows in north of Iran with lower error than the (ARIMA) model.
Authors and Affiliations
Mahdi Moharrampour - Abdulhamid mehrabi - hooman Hajikandi - Saeed Sohrabi - Javad Vakili
Keywords
Related Articles
- EP ID EP32944
- DOI -
- Views 526
- Downloads 0
Comparative Study of Canal Design
India is basically an agrarian country and crop-water is supplied through the system of canals from either dams or distribution chambers. Conventionally, canals are designed, constructed & maintained by the Water Reso...
Comparison of Support Vector Machines (SVM) and Autoregressive integrated moving average (ARIMA) in daily flow forecasting
In recent years, Computational Intelligence (CI) is applied to solve problems for some physical processes with nonlinear relations. Use of data, extraction of relations between them and generalizing in other situation...
Observations and Modeling of Transverse Mixing in a Natural Gravel-Bed River
A comprehensive, continuous, point-source, dye concentration experiment was performed on the Kootenai River, Idaho, USA to describe natural river transport and mixing dynamics. Centerline fluorescent dye concentration...
Investigation of Particle-Size Distribution and Friction Factor for a Gravel-Bed river: Marbar
The study of flow and sediment size distribution in gravel –bed Rivers has been the pivotal attention of many researchers for more than three decades. The surface grain size of gravel and cobble-bed rivers reflects the ca...
Bed Shear Stress around a Slender Obstruction in Different Channel Section
Study of shear stress in bed channel may be considered as prerequisite for study of sediment transportation, as sediment transports due to shear stress which acts because of fluid flowing above it. Shear stress in bed...