The advancement of the machine learning (ML) models has demonstrated notable progress in geosciences. They can identify the underlying process or causality of natural hazards. This article introduces the development and verification procedures of a new hybrid ML model, namely Bat-ELM for predictive drought modelling. The multi-temporal standardized precipitation evapotranspiration index (SPEI-3 and SPEI-6) is computed as the meteorological drought index at two study regions (Beypazari and Nallihan), located in Ankara province, Turkey. The proposed hybrid model is obtained by integrating the Bat optimization algorithm as the parameter optimizer with an extreme learning machine (ELM) as the regressor engine. The efficiency of the intended model was evaluated against the classic artificial neural network (ANN) and standalone ELM models. The evaluation and assessment are conducted using statistical metrics and graphical diagrams. The forecasting results showed that the accuracy of the proposed model outperformed the benchmark models. In a quantitative assessment, the Bat-ELM model attained minimal root mean square error for the SPEI-3 and SPEI-6 (RMSE=0.58 and 0.43 at Beypazari station and RMSE=0.53 and 0.37 at Nallihan station) over the testing phase. This indicates the new model approximately 20 and 15% improves the forecasting accuracy of traditional ANN and classic ELM techniques, respectively.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
This paper compares the classification and prediction capabilities of decision tree (DT), genetic programming (GP), and gradient boosting decision tree (GBT) techniques for one-month ahead prediction of standardized precipitation index in Ankara province and standardized precipitation evaporation index in central Antalya region. The evolved models were developed based on multi-station prediction scenarios in which observed (reanalyzed) data from nearby stations (grid points) were used to predict drought conditions in a target location. To tackle the rare occurrence of extreme dry/wet conditions, the drought series at the target location was categorized into three classes of wet, normal, and dry events. The new models were trained and validated using the frst 70% and last 30% of the datasets, respectively. The results demonstrated the promising performance of GBT for meteorological drought classification. It provides better performance than DT and GP in Ankara; however, GP predictions for Antalya were more accurate in the testing period. The results also exhibited that the proposed GP model with a scaled sigmoid function at root can efortlessly classify and predict the number of dry, normal, and wet events in both case studies.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.