Multi resolution genetic programming approach for stream flow forecasting

Authors:
Rathinasamy Maheswaran;Rakesh Khosa
Affiliations:
Department of Civil Engineering, Indian Institute of Technology Delhi, New Delhi, India;Department of Civil Engineering, Indian Institute of Technology Delhi, New Delhi, India
Venue:
SEMCCO'11 Proceedings of the Second international conference on Swarm, Evolutionary, and Memetic Computing - Volume Part I
Year:
2011

Citing 5
Cited 0

Genetic programming: an introduction: on the automatic evolution of computer programs and its applications

Genetic programming: an introduction: on the automatic evolution of computer programs and its applications
Quantum computing applications of genetic programming

Advances in genetic programming
Prediction and Modelling of the Flow of a Typical Urban Basin through Genetic Programming

Proceedings of the Applications of Evolutionary Computing on EvoWorkshops 2002: EvoCOP, EvoIASP, EvoSTIM/EvoPLAN
Detection of Incidents on Motorways in Low Flow High Speed Conditions by Genetic Programming

Proceedings of the Applications of Evolutionary Computing on EvoWorkshops 2002: EvoCOP, EvoIASP, EvoSTIM/EvoPLAN
Sea water level forecasting using genetic programming and comparing the performance with Artificial Neural Networks

Computers & Geosciences

Quantified Score

Hi-index	0.00

Visualization

Abstract

Genetic Programming (GP) is increasingly used as an alternative for Artificial Neural Networks (ANN) in many applications viz. forecasting, classification etc. However, GP models are limited in scope as their application is restricted to stationary systems. This study proposes use of Multi Resolution Genetic Programming (MRGP) based approach as an alternative modelling strategy to treat non-stationaries. The proposed approach is a synthesis of Wavelets based Multi-Resolution Decomposition and Genetic Programming. Wavelet transform is used to decompose the time series at different scales of resolution so that the underlying temporal structures of the original time series become more tractable. Further, Genetic Programming is then applied to capture the underlying process through evolutionary algorithms. In the case study investigated, the MRGP is applied for forecasting one month ahead stream flow in Fraser River, Canada, and its performance compared with the conventional, but scale insensitive, GP model. The results show the MRGP as a promising approach for flow forecasting.