Evolutionary Polynomial Regression application for missing data handling in meteo-climatic gauging stations

One of the most often encountered modelling problems is that of handling missing data, i.e. the problem of intermediate data gaps, where data/observations before and after the missing observations are available. The gaps in data represent discontinuities, which can pose difficulties both for model construction and model application phases. Evolutionary Polynomial Regression (EPR-MOGA) is a data-driven hybrid technique, which combines the effectiveness of genetic programming with the numerical regression for developing simple and easily interpretable mathematical model expressions. Evolutionary Polynomial Regression takes advantage of the evolutionary computing approach that allows the construction of several model expressions based on training data and least squares methodology to estimate numerical parameters/coefficients. These models can then be verified on a test set and gaps can be in-filled in test datasets by using one selected model. Because of the pseudo-polynomial formulations achievable by EPR-MOGA, it requires fewer numbers of parameters to be estimated, which in turn requires shorter time series for training. Another advantage of the EPR-MOGA approach is the ability to choose objective functions pertaining accuracy and parsimony. In the present work, an application of EPR-MOGA is shown on some sites belonging to the Apulian meteo-climatic monitoring network.