Search

Accurate streamflow prediction is of great significance for water resource management. In recent years, data-driven models such as artificial neural networks (ANNs) and support vector machines (SVMs) have been widely used in the field of flow prediction. However, traditional data-driven models neglect the extraction and utilization of the data's own characteristics. This study proposes a daily flow prediction model based on the pattern recognition of flow sequences. Based on the input number of the prediction model derived from the partial autocorrelation function, the flow sequence was divided into subsequences. Five patterns of flow subsequences, including monotonic rising, monotonic falling, monotonic stable, concave, and convex, were then identified, which helped to explore the characteristics of the flow subsequences. For each pattern, traditional ANN and SVM models were applied to predict the flow. A comparison with the traditional ANN and SVM models shows that the hybrid models of the pattern recognition method (PRM) and the traditional ANN and SVM have higher accuracy. The Nash efficiency coefficient (NSE) of the hybrid PRM-SVM model was as high as 0.9815, and the mean absolute percentage error (MAPE) was only 6.75%. In addition, the prediction accuracy of the flood peak also improved. The average relative error of the peak flood derived from the hybrid PRM-ANN and PRM-SVM models were reduced by 0.12% and 0.40%, respectively, compared with the traditional ANN and SVM models. [ABSTRACT FROM AUTHOR]

Precise and reliable long-term streamflow prediction contributes to water resources planning and management. Artificial neural network (ANN) have shown its remarkable ability in forecasting non-linear hydrological processes without involvement of complex, dynamic, hydrological and hydro-climatologic physical process in the water shed. To improve its non-stationary responses, decomposition methods are adopted as pre-processing methods in this study including Empirical Mode Decomposition (EMD), Ensemble Empirical Mode Decomposition (EEMD) and Seasonal-Trend decomposition using Loess (STL). The original time sequence is decomposed to several components, which are then taken as the inputs of the ANN model. EMD and EEMD are data- adaptable methods, and thus the number of Intrinsic Mode Functions (IMFs) might differ for different sequences, leading to the discrepancy of the input number for ANN model in training and predicting. Fisher's ordered clustering is thus used to classify the IMFs into a determined number of classes based on their frequency spectrum resulting from Maximum Entropy Spectral Analysis (MESA). The proposed methodology is applied on four important hydrological stations on the upper stream of the Yellow River and the Yangtze River in China, respectively, to forecast the streamflow of the next whole year with the historical daily data of the past 6 years. The Nash-Sutcliffe efficiencies of the monthly prediction are higher than 0.85 for all of the four cases, and various indicators indicates that the proposed hybrid method of STL-ANN performs better than other compared methods. The highlights of this study lies in that only historical daily streamflow data is used to derive an accurate long-term prediction by data mining based on decomposition technology and mapping relationships between the decomposed components and the original sequence in the future. [ABSTRACT FROM AUTHOR]

Artificial neural network (ANN) combined with signal decomposing methods is effective for long‐term streamflow time series forecasting. ANN is a kind of machine learning method utilized widely for streamflow time series, and which performs well in forecasting nonstationary time series without the need of physical analysis for complex and dynamic hydrological processes. Most studies take multiple factors determining the streamflow as inputs such as rainfall. In this study, a long‐term streamflow forecasting model depending only on the historical streamflow data is proposed. Various preprocessing techniques, including empirical mode decomposition (EMD), ensemble empirical mode decomposition (EEMD) and discrete wavelet transform (DWT), are first used to decompose the streamflow time series into simple components with different timescale characteristics, and the relation between these components and the original streamflow at the next time step is analyzed by ANN. Hybrid models EMD‐ANN, EEMD‐ANN and DWT‐ANN are developed in this study for long‐term daily streamflow forecasting, and performance measures root mean square error (RMSE), mean absolute percentage error (MAPE) and Nash–Sutcliffe efficiency (NSE) indicate that the proposed EEMD‐ANN method performs better than EMD‐ANN and DWT‐ANN models, especially in high flow forecasting. [ABSTRACT FROM AUTHOR]

According to diagnostic criteria, skin tumors can be divided into three categories: benign, low degree and high degree malignancy. For high degree malignant skin tumors, if not detected in time, they can do serious harm to patients' health. However, in clinical practice, identifying malignant degree requires biopsy and pathological examination which is time costly. Furthermore, in many areas, due to the severe shortage of dermatologists, it's inconvenient for patients to go to hospital for examination. Therefore, an easy to access screening method of malignant skin tumors is needed urgently. Firstly, we spend 5 years to build a dataset which includes 4,500 images of 10 kinds of skin tumors. All instances are verified pathologically thus trustworthy; Secondly, we label each instance to be either low-risk, high-risk or dangerous in which Junctional nevus, Intradermal nevus, Dermatofibroma, Lipoma and Seborrheic keratosis are low-risk, Basal cell carcinoma, Bowen's disease and Actinic keratosis are high-risk, Squamous cell carcinoma and Malignant melanoma are dangerous; Thirdly, we apply the Xception architecture to build the risk degree classifier. The area under the curve (AUC) for three risk degrees reach 0.959, 0.919 and 0.947 respectively. To further evaluate the validity of the proposed risk degree classifier, we conduct a competition with 20 professional dermatologists. The results showed the proposed classifier outperforms dermatologists. Our system is helpful to patients in preliminary screening. It can identify the patients who are at risk and alert them to go to hospital for further examination. [ABSTRACT FROM AUTHOR]

Solar radiation prediction is significant for solar energy utilization. This paper presents hybrid methods following the decomposition-prediction-reconfiguration paradigm using only historical radiation records with different combination of decomposition methods, Ensemble Empirical Mode Decomposition (EEMD) and Wavelet Analysis (WA), and the reconfiguration methods, regression model (RE) and Artificial Neural Network (ANN). The application in west China indicates that these hybrid decomposition-reconfiguration models perform well for monthly prediction, while the comparisons of the daily prediction show that the hybrid EEMD-RE model has a higher degree of fitting and a better prediction effect in long-term prediction of solar radiation intensity, which verifies (1) decomposition of original solar radiation data results in components with regular characteristics; (2) the relationship between the original solar radiation sequence and the derived intrinsic mode functions (IMFs) is linear; and (3) EEMD has strong adaptivity for non-linear and non-stationary series. The proposed hybrid decomposition-reconfiguration models have great application prospect for monthly long-term prediction of solar radiation intensity, especially in the areas where complex climate data is difficult to obtain, and the EEMD-RE model is recommended for the daily long-term prediction. [ABSTRACT FROM AUTHOR]

Long-term prediction of solar radiation intensity plays an important role in the planning and design of photovoltaic power stations. Unlike previous research on solar radiation prediction requiring various meteorological and topographic data, this study proposed a rolling prediction model combining Empirical Mode Decomposition (EMD) and Artificial Neural Network (ANN) techniques with the need for historical solar radiation data only. To overcome the inconsistency problem of the number of intrinsic mode functions derived from the EMD, they are classified into high-frequency term, low-frequency term, and trend item, which are taken as the input parameters of the ANN model. With the historical data one-year after as the output parameter, the ANN model implies the complex, non-linear relationship between the adjacent periods, and it can be used to predict long-term solar radiation. The proposed methodology is applied to Gonghe county in the Qinghai province of China, where a large-scale photovoltaic power plant is under planning. The results indicate that the correlation coefficients between the daily and monthly predicted value and the historical data are 0.698 and 0.930, respectively, which are comparable to previous studies with a greater data requirement and a simpler model. [ABSTRACT FROM AUTHOR]