Model Data Selection and Data Pre-processing Approaches

摘要: Data-based modeling relies on historical data without directly taking account of underlying physical processes in hydrology . So, real-world hydrological commonly requires a complex input structure and very lengthy training to represent inherent dynamic systems. In cases where large amount is available, all which used for modeling, technical issues such as the increase computational complexity lack memory spaces have been observed. The likelihood these problems occurring much greater case models possess high nonlinearity number parameters. Therefore, there definite need identify proper techniques adequately reduce inputs required length nonlinear models. Removing redundant from available pools deciding upon optimum make reliable prediction are main purposes approaches. This section book describes abilities novel Gamma Test (GT), entropy theory (ET), Principle Component Analysis (PCA), cluster analysis (CA), Akaike’s Information Criterion (AIC ), Bayesian (BIC ) model selection. novelty this work that many approaches first time scenarios solar radiation estimation, rainfall-runoff , evapotranspiration Towards end chapter, conventional selection procedures Cross-Correlation Approach (CCA), Cross-Validation (CVA), Data Splitting (DSA) explained detail. These traditional were check authenticity newly applied methods later study chapters.