Search

In the recent times, the use of geosynthetic-reinforced soil (GRS) technology has become popular for constructing safe and sustainable pavement structures. The strength of the subgrade soil is routinely assessed in terms of its California bearing ratio (CBR). However, in the past, no effort was made to develop a method for evaluating the CBR of the reinforced subgrade soil. The main aim of this paper is to explore and appraise the competency of the several intelligent models such as artificial neural network (ANN), least median of squares regression, Gaussian processes regression, elastic net regularisation regression, lazy K-star, M-5 model trees, alternating model trees and random forest in estimating the CBR of reinforced soil. For this, all the models were calibrated and validated using the reliable pertinent historical data. The prognostic veracity of all the tools mentioned supra were assessed using the well-established traditional statistical indices, external model evaluation technique, multi-criteria assessment approach and independent experimental dataset. Due to the overall excellent performance of ANN, the model was converted into a trackable functional relationship to estimate the CBR of reinforced soil. Finally, the sensitivity analysis was performed to find the strength and relationship of the used parameters on the CBR value. [ABSTRACT FROM AUTHOR]

Wildfires are natural or man‐made disasters that continuously threaten portions of the transmission and distribution grid, and thus the stability of the electric grid. This paper presents a two‐stage framework for assessing power system‐wildfire risk using a data‐driven wildfire prediction model. The first stage of the framework estimates the spatio‐temporal probability of potential wildfire ignition and propagation using a deep neural network in combination with the wildfire physical spread model. Analysis reveals similar spatial and temporal patterns between the model‐predicted wildfire ignition potential and actual wildfire ignition. Motivated by these observations, the second stage assesses the wildfire risk in the power grid operation in terms of potential loss of load by de‐energisation, through combining geospatial information system data of the power grid topology and the stochastic spatio‐temporal wildfire model developed in the first stage. The electric power utility applications introduced by the proposed framework are twofold: 1) a spatio‐temporal risk model for proactive de‐energisation against potential power system failure‐induced wildfire, and 2) a spatio‐temporal spreading model for optimal grid operations against exogenous wildfire. The proposed model, based on real‐world dataset, is demonstrated on the IEEE 24‐bus test system mapped to a study area in Northern California, while the results illustrate the proposed model can achieve the best performance in potential wildfire ignition detection (AUC of 0.995) compared to other baselines, as well as demonstrates the risk‐aware operation of the power system enabled by the proposed framework. [ABSTRACT FROM AUTHOR]

Recorded time series of relative humidity (RH) are modeled by using genetic expression programming (GEP) and artificial neural networks (ANNs) models. The data are noisy and contain missing datapoints. RH is modeled as a function of three meteorological variables: temperature, wind speed, and pressure. Various model structures of both of these models are investigated with the aim of testing the robustness of the predicted values in the presence of noise and missing data. Due to the presence of noise, a sophisticated treatment of missing data was not justifiable, and therefore, the strategy adopted was just to carry the datapoints backward, although this may induce bias in the time dimension and contaminate the predicted results. The results of this study indicate that through a careful selection of model structures both GEP and ANN can produce adequately reliable prediction of RH values 1 year into the future. The paper provides evidence that this model structure is feasible when the dependent variables include both the present and past values. [ABSTRACT FROM AUTHOR]

Wildfires have caused devastating consequences to property and human and animal lives, which has become a global problem. Consequently, advanced wildfire prediction models are required to treat complex features and climate conditions. As a result, Machine Learning and Deep Learning models are becoming popular. However, creating a balanced true and false labeled dataset in the wild-fire domain is often challenging. Hence, One-class classification models are a promising approach to overcome this concern. In this paper, several One-class classification models are investigated; linear models (Principal Component Analysis and One-Class Support Vector Machines), outlier ensemble models (Lightweight On-line Detector of Anomalies and Locally Selective Combination of Parallel Outlier Ensembles), proximity-based models (Histogram-based Outlier Score and Rotation-based Outlier Detection), probabilistic models (Unsupervised Outlier Detection Using Empirical Cumulative Distribution Functions and Copula-Based Outlier Detection), neural network-based models (Deep One-class Classification and Adversarially Learned Anomaly Detection) is used in two case studies for California and Western Australian states. In conclusion, it was found that Deep learning-based One-class classification models outperform other models in terms of performance and feature importance of showcasing the effectiveness of deep neural network models in the wildfire prediction domain. [ABSTRACT FROM AUTHOR]

This paper evaluates the potential of five modeling approaches, namely M5 model tree, random forest, artificial neural networks, support vector machines and Gaussian processes, for the prediction of unconfined compressive strength of stabilized pond ashes with lime and lime sludge. The study not only presents five models for the same set of data but also compares the overall performance of them. Dataset used consists of 255 samples acquired from laboratory experiments. Out of the total, 170 randomly chosen samples were used for training and remaining 85 were used for testing the models. Input dataset consists of eight parameters (uniformity coefficient, coefficient of curvature, maximum dry density, optimum moisture content, lime, lime sludge, curing period and 7-day soaked California bearing ratio), while the output is UCS value at 7, 28, 45, 90 and 180 days of curing. Comparisons of results propose that Gaussian processes modeling strategy works well and the overall performance was substantially nearer to the exact agreement line. As a result of GP model, higher value of CC = 0.997 and lower values of RMSE = 23.016 kPa and MAE = 16.455 were obtained for testing the dataset. Sensitivity analysis suggests that lime, lime sludge, curing period and California bearing ratio are the significant parameters for predicting the unconfined compressive strength of stabilized pond ashes. The results confirmed that GP models are in a position to predict the unconfined compressive strength of stabilized pond ashes with an excessive degree of accuracy; however, GP modeling approach proves that this approach is more economical and less difficult in comparison with tedious laboratory work. [ABSTRACT FROM AUTHOR]

California bearing ratio (CBR) is an important property used to express the quality and strength of the unbound granular materials and subgrade soils. It is one of the material inputs for the American Association State Highway Transportation Officials 1993 guide, and the Mechanistic Empirical Pavement Design Guide for the structural design of flexible pavements in case of the resilient modulus is not known. CBR is also conducted on the unbound materials for the quality control/quality assurance during construction. Because of its importance, this paper presents an attempt to develop simple and reliable CBR models based on routine material properties such as gradation, Atterberg limits and compaction properties using regression analysis (RA) and artificial neural networks (ANNs). Database of 207 CBR values was collected from the quality control reports prepared at the Highway and Airport Engineering Laboratory, Mansoura University. The collected CBR values were found to range between 26 and 98%. About 80% of the collected data were used for model development, while the remaining 20% were used for model validation in addition to 11 laboratory tested specimens. The developed model by RA and ANNs correlates CBR values with maximum dry density and diameter at 60% passing (D60). The prediction accuracy in terms of coefficient of determination ( R 2 ) for the developed CBR model by both techniques was excellent, and the validation of the suggested model was satisfactory. [ABSTRACT FROM AUTHOR]

Cardiovascular Disease (CVD) is one of the most catastrophic and life threatening health issue nowadays. Early detection of CVD is an important solution to reduce its devastating effects on health. In this paper, an efficient CVD detection algorithm is identified. The algorithm uses patient demographic data as inputs, along with several ECG signal features extracted automatically through signal processing techniques. Cross-validation results show a 98.29 % accuracy for the decision tree classification algorithm. The algorithm has been integrated into a web based system that can be used at anytime by patients to check their heart health status. At one end of the system is the ECG sensor attached to the patient's body, while at the other end is the detection algorithm. Communication between the two ends is done through an Android application. [ABSTRACT FROM AUTHOR]

The application of expert systems has emerged as a cost efficient tool in civil engineering over the past decades. Use of these tools for decision making in land use will be outlined in this paper. In this study, soil-structure interaction (SSI) is discussed by using expert systems, namely, neural network (NN) approaches. This method provides a new point of view for evaluations of SSI and land use. Data from 58 local sites in California, namely, earthquake, structure, and soil property data, are used. In the expert system approach, two NN architectures are used: the back propagation NN architecture and the general regression NN (GRNN) architecture. There are 21 parameters considered as input and 4 output parameters. The four output parameters chosen are soil-to-structure rigidity ratio, period lengthening, foundation damping, and whether SSI effects can be neglected or not. The results show that the GRNN approaches are more useful and powerful for evaluation of SSI and land use. [ABSTRACT FROM AUTHOR]

An accurate and efficient Short-Term Load Forecasting (STLF) plays a vital role for economic operational planning of both regulated and restructured power systems. This paper presents the STLF models by introducing wavelet transforms, in different ways, with feed-forward Neural Networks (NNs). First, a wavelet-based NN is modeled, where the forecasting has been accomplished in three-stages and the wavelet technique is employed to decompose/reconstruct the original signals and non-linearity of the decomposed signals. Second, an Adaptive Wavelet Neural Network (AWNN) is modeled, which is a new class of NN with continuous wavelet function as the hidden layer node's activation function. Unlike the first model, AWNN does not externally decompose/ reconstruct the original signals and, therefore, this model deals with the problem related to loss of high frequency information that might occur in the wavelet-based NN model. AWNN continuously updates the wavelet parameters (translation and dilation) and layer weights through a back-propagation training algorithm as in classical NNs. The performances of these two models are compared with Multi-Layer Perceptron NN (MLPNN) with the application of day-ahead and hour-ahead load forecasting in the California electricity market. The results are also compared with California Independent System Operator (CAISO)'s forecasted system loads. It is found that due to faster and accurate training capability, AWNN outperforms the MLPNN, wavelet-based NN and CAISO load forecasts. [ABSTRACT FROM AUTHOR]

The generalization error bounds found by current error models using the number of effective parameters of a classifier and the number of training samples are usually very loose. These bounds are intended for the entire input space. However, support vector machine (SVM), radial basis function neural network (RBFNN), and multilayer perceptron neural network (MLPNN) are local learning machines for solving problems and treat unseen samples near the training samples to be more important. In this paper, we propose a localized generalization error model which bounds from above the generalization error within a neighborhood of the training samples using stochastic sensitivity measure. It is then used to develop an architecture selection technique for a classifier with maximal coverage of unseen samples by specifying a generalization error threshold. Experiments using 17 University of California at Irvine (UCI) data sets show that, in comparison with cross validation (CV), sequential learning, and two other ad hoc methods, our technique consistently yields the best testing classification accuracy with fewer hidden neurons and less training time. [ABSTRACT FROM AUTHOR]

The current study aims at predicting the strength of the problematic clayey soils treated with combinations of pozzolan of natural sources and lime powder when added as soil additives at a nano scale. Multiple linear regression (MLR), artificial neural networks (ANN) and fuzzy logic (FL) tools were employed in the analytical study. The variables of the present study include the following: nano pozzoaln of natural source (NNP) content, nano lime content (NL), median particle size of NNP, active silica content of NNP (SiO2active), Initial liquid limit (ILL) and initial plastic limit (IPL) of the investigated soils. NNP was added at five percentages, i.e. 0%, 0.5%, 1%, 1.5% and 2%, while NL was added at five percentages, i.e. 0%, 0.3%, 0.6%, 0.9% and 1.2%. Three median particle sizes namely 50, 100 and 500 nm size were studied. Based on the different investigated soils and combinations, 120 soil mixtures were prepared and tested. California bearing ratio (CBR) and plasticity index (PI) were particularly examined. CBR tests were conducted at a soaked condition on specimens compacted to a maximum dry density (MDD) at the optimum moisture content (OMC). PI values were obtained following the Atterberg limits test. Based on the results of the performance criteria of the developed predictive models, it can be concluded that the CBR and PI of the expansive clayey soils can be effectively predicted using ANN and FL techniques. The results obtained by MLR were far from those obtained by both ANN & FL. In addition, ANN tool was slightly more accurate than FL as far as prediction of CBR and PI is concerned. The higher capability of ANN & FL models in predicting CBR & PI values, which generally obtained through time-consuming and expensive tests, could be useful for geotechnical engineers to assess or design a new pavement project. Further, it is recommended to do a re-evaluation of the current study in future, particularly when more data is available in the literature. [ABSTRACT FROM AUTHOR]

The increasing incidence and severity of wildfires underscores the necessity of accurately predicting their behavior. While high-fidelity models derived from first principles offer physical accuracy, they are too computationally expensive for use in real-time fire response. Low-fidelity models sacrifice some physical accuracy and generalizability via the integration of empirical measurements, but enable real-time simulations for operational use in fire response. Machine learning techniques have demonstrated the ability to bridge these objectives by learning first-principles physics while achieving computational speedups. While deep learning approaches have demonstrated the ability to predict wildfire propagation over large time periods, time-resolved fire-spread predictions are needed for active fire management. In this work, we evaluate the ability of deep learning approaches in accurately modeling the time-resolved dynamics of wildfires. We use an autoregressive process in which a convolutional recurrent deep learning model makes predictions that propagate a wildfire over 15 min increments. We apply the model to four simulated datasets of increasing complexity, containing both field fires with homogeneous fuel distribution as well as real-world topologies sampled from the California region of the United States. We show that even after 100 autoregressive predictions representing more than 24 h of simulated fire spread, the resulting models generate stable and realistic propagation dynamics, achieving a Jaccard score between 0.89 and 0.94 when predicting the resulting fire scar. The inference time of the deep learning models are examined and compared, and directions for future work are discussed. [ABSTRACT FROM AUTHOR]

Purpose – The purpose of this paper is to present an artificial intelligent (AI) system for estimating the construction cost of water and sewer rehabilitation projects. Design/methodology/approach – To develop the proposed system, data pertaining to 54 sewer and water rehabilitation projects was collected. The collected data were analyzed using Pareto analysis technique to identify the most important factors that contribute positively to the cost estimation process. These factors were then utilized to develop a neural network (NN) model that estimates the construction cost of this class of projects. Findings – The study reveals a set of 23 factors that highly impact the construction cost of water and sewer network rehabilitation projects and presents a NN model that predicts the cost of these projects with high accuracy. Research limitations/implications – The proposed system was developed using information obtained from the city of San Diego, California, USA. The cost of these projects ranged from $800,000 to $7 million. The diameter of pipes installed in these projects ranged from 1 in. to 36 in. and their length was up to about 2.7 miles. Originality/value – The developed system saves time, improves the accuracy of the estimates and prevents problems that are usually associated with inaccurate estimates. The system will not only help funding authorities to ensure maximum utilization of resources, but will also help cities to manage their expenditures in a manner that assures satisfactory performance of their buried assets. Furthermore, the developed system is also believed to assist cities in comparing alternatives and the go/no-go decision making process. [ABSTRACT FROM AUTHOR]

The sediment load transported in a river is the most complex hydrological phenomenon due to a large number of obscure parameters and the existence of both spatial variability of the basin characteristics and temporal climatic patterns. In this paper two artificial neural network (ANN) models were developed for semidistributed modeling of the suspended sediment load process of the Eel River watershed located in California. The first model was an integrated ANN model trained by the data of multiple stations inside the watershed. In the second model, a geomorphology-based ANN model, space-dependent geomorphologic parameters of the subbasins, extracted by geographic information system tools, accompanied by time-dependent meteorological data, were imposed on the network. In both models, three-layer perceptron neural networks were trained considering various combinations of input and hidden layers' neurons, and the optimum architectures of the models were selected according to the computed evaluation criteria. Furthermore, the ability of the models for spatiotemporal modeling of the process was examined through the cross-validation technique for a station. The obtained results demonstrate that although the predicted sediment load time series by both models are in satisfactory agreement with the observed data, the geomorphological ANN model produces better performance than an integrated model because it employs spatially variable factors of the subbasins as the model's inputs. Therefore, the model can operate as a nonlinear time-space regression tool rather than a fully lumped model. [ABSTRACT FROM AUTHOR]

Predictive control has gained increasing attention for its ability to reduce energy consumption and improve occupant comfort in buildings. The plug loads prediction is a key component for the predictive building controls, as plug loads is a major source of internal heat gains in buildings. This study proposed a novel method to apply the Long-Short-Term-Memory (LSTM) Network, a special form of Recurrent Neural Network, to predict plug loads. The occupant count and the time have been confirmed to drive the plug load profile and thus selected as the features for the plug load prediction. The LSTM network was trained and tested with ground truth occupant count data collected from a real office building in Berkeley, California. Results from the LSTM network markedly improve the prediction accuracy compared with traditional linear regression methods and the classical Artificial Neural Network. 95% of 1-h predictions from LSTM network are within ±1 kW of the actual plug loads, given the average plug loads during the office hour is 8.6 kW. The CV(RMSE) of the predicted plug load is 11% for the next hour, and 20% for the next 8 h. Lastly, we compared four prediction approaches with the office building we monitored: LSTM vs. ARIMA, with occupant counts vs. without occupant counts. It was found, the prediction error of the LSTM approach is around 4% less than the ARIMA approach. Using occupant counts as an exogenous input could further reduce the prediction error by 5%–6%. The findings of this paper could shed light on the plug load prediction for building control optimizations such as model-predictive control. • Accurate plug loads prediction is key to Model Predictive Control in buildings. • LSTM (Long Short Tem Memory) network is used to predict plug loads using occupant count data. • The prediction accuracy is 89% for the next hour, and 80% for the next 8 h. • LSTM is computational demanding but 4% more accurate than ARIMA. • Using occupant counts could reduce plug load prediction error by 5%–6%. [ABSTRACT FROM AUTHOR]

This paper proposes a multiclassification algorithm using multilayer perceptron neural network models. It tries to boost two conflicting main objectives of multiclassifiers: a high correct classification rate level and a high classification rate for each class. This last objective is not usually optimized in classification, but is considered here given the need to obtain high precision in each class in real problems. To solve this machine learning problem, we use a Pareto-based multiobjective optimization methodology based on a memetic evolutionary algorithm. We consider a memetic Pareto evolutionary approach based on the NSGA2 evolutionary algorithm (MPENSGA2). Once the Pareto front is built, two strategies or automatic individual selection are used: the best model in accuracy and the best model in sensitivity (extremes in the Pareto front). These methodologies are applied to solve 17 classification benchmark problems obtained from the University of California at Irvine (UCI) repository and one complex real classification problem. The models obtained show high accuracy and a high classification rate for each class. [ABSTRACT FROM AUTHOR]

Evapotranspiration is one of the basic components of the hydrologic cycle and essential for estimating irrigation water requirements. This paper investigates the modelling of evapotranspiration using the feed-forward artificial neural network (ANN) technique with the Levenberg--Marquardt (LM) training algorithm. The LM algorithm has never been used in evapotranspiration estimation before. The LM is used for the optimization of network weights, since this algorithm is more powerful and faster than the conventional gradient descent. Various combinations of daily climatic data, i.e. wind speed, air temperature, relative humidity and solar radiation, from three stations in Los Angeles, USA, are used as inputs to the ANN so as to evaluate the degree of effect of each of these variables on evapotranspiration. A comparison is made between the estimates provided by the ANN and those of the following empirical models: Penman, Hargreaves, Turc. Mean square error, mean absolute error and determination coefficient statistics are used as comparing criteria for the evaluation of the models' performances. Based on the comparisons, it was found that the neural computing technique could be employed successfully in modelling evapotranspiration process from the available climatic data. The results also indicate that the Hargreaves method provides better performance than the Penman and Turc methods in estimation of the evapotranspiration. The accuracy of the ANN technique in evapotranspiration estimation using nearby station data was also investigated. [ABSTRACT FROM AUTHOR]

Recent researches are directed towards the regional seismic risk assessment of structures based on a bridge inventory analysis. The framework for traditional regional risk assessments consists of grouping the bridge classes and generating fragility relationships for each bridge class. However, identifying the bridge attributes that dictate the statistically different performances of bridges is often challenging. These attributes also vary depending on the demand parameter under consideration. This paper suggests a multi-parameter fragility methodology using artificial neural network to generate bridge-specific fragility curves without grouping the bridge classes. The proposed methodology helps identify the relative importance of each uncertain parameter on the fragility curves. Results from the case study of skewed box-girder bridges reveal that the ground motion intensity measure, span length, and column longitudinal reinforcement ratio have a significant influence on the seismic fragility of this bridge class. [ABSTRACT FROM AUTHOR]

The objective of the research is to estimate the value of the California bearing ratio (CBR) through the application of ANN. The methodology consists of creating a database with soil index and CBR variables of the subgrades and granular base of pavements in Jaen, Peru, carried out in the soil mechanics laboratories of the city and the National University of Jaen. In addition, the Python library Seaborn is for variable selection and relevance, and the scikit-learn and Keras libraries were used for the learning, training, and validation stage. Five ANN are proposed to estimate the CBR value, obtaining an error of 4.47% in the validation stage. It can be concluded that this method is effective and valid to determine the CBR value in subgrades and granular bases of any pavement for its evaluation or design. [ABSTRACT FROM AUTHOR]

This paper reports the refinement of a published empirical model of salinity in the San Francisco Bay-Delta estuary by integration with a Bayesian artificial neural network (ANN) model and incorporation of additional inputs. Performance goals established for the resulting hybrid model are based on the quality of fit to observed data (replicative and predictive validation) as well as sensitivity when compared with a priori knowledge of system behavior (structural validation). ANN model parameters were constrained to provide plausible sensitivity to coastal water level, a key input introduced in the hybrid formulation. In addition to representing observed data better than the underlying empirical model while meeting structural validation goals, the hybrid model allows for characterization of prediction uncertainty. This work demonstrates a real-world application of a general approach— integration of a pre-existing model with a Bayesian ANN constrained by knowledge of system behavior—that has broad application for environmental modeling. [ABSTRACT FROM AUTHOR]

Salinity management is a subject of particular interest in estuarine environments because of the underlying biological significance of salinity and its variations in time and space. The foremost step in such management practices is understanding the spatial and temporal variations of salinity and the principal drivers of these variations. This has traditionally been achieved with the assistance of empirical or process-based models, but these can be computationally expensive for complex environmental systems. Model emulation based on data-driven methods offers a viable alternative to traditional modeling in terms of computational efficiency and improving accuracy by recognizing patterns and processes that are overlooked or underrepresented (or overrepresented) by traditional models. This paper presents a case study of emulating a process-based boundary salinity generator via deep learning for the Sacramento–San Joaquin Delta (Delta), an estuarine environment with significant economic, ecological, and social value on the Pacific coast of northern California, United States. Specifically, the study proposes a range of neural network models: (a) multilayer perceptron, (b) long short-term memory network, and (c) convolutional neural network-based models in estimating the downstream boundary salinity of the Delta on a daily time-step. These neural network models are trained and validated using half of the dataset from water year 1991 to 2002. They are then evaluated for performance in the remaining record period from water year 2003 to 2014 against the process-based boundary salinity generation model across different ranges of salinity in different types of water years. The results indicate that deep learning neural networks provide competitive or superior results compared with the process-based model, particularly when the output of the latter are incorporated as an input to the former. The improvements are generally more noticeable during extreme (i.e., wet, dry, and critical) years rather than in near-normal (i.e., above-normal and below-normal) years and during low and medium ranges of salinity rather than high range salinity. Overall, this study indicates that deep learning approaches have the potential to supplement the current practices in estimating salinity at the downstream boundary and other locations across the Delta, and thus guide real-time operations and long-term planning activities in the Delta. [ABSTRACT FROM AUTHOR]

• This paper compared Intelligent Genetic Algorithm to ANN in investigating fixed object crashes among elderly drivers. • Comparing to ANN, Intelligent Genetic Algorithm was more capable of predicting high-severity crashes. • ANN was more accurate in predicting low-severity crashes, while it was not able to detect more severe crashes. • Too much complexity did not only create over-fitting problems but it also increased the model run-time significantly. • The light condition was identified as the most significant factor, followed by the existence of right and left shoulders. Run-off-road (ROR) crashes have always been a major concern as this type of crash is usually associated with a considerable number of serious injury and fatal crashes. A substantial portion of ROR fatalities occur in collisions with fixed objects at the roadside. Thus, this study seeks to investigate the severity of ROR crashes where elderly drivers, aged 65 years or more, hit a fixed object. The reason why the present study investigates this issue among older drivers is that, comparing to younger drivers, this age group of drivers have different psychological and physical features. Because of these differences, they are more likely to get injured in ROR types of crashes. This paper applies two types of Artificial Intelligence (AI) techniques, including hybrid Intelligent Genetic Algorithm and Artificial Neural Network (ANN) using the crashe information of California in 2012 obtained from Highway Safety Information System (HSIS) database. Although the results showed that the developed ANN outperformed the hybrid Intelligent Genetic Algorithm, the hybrid approach was more capable of predicting high-severity crashes. This is rooted in the way the hybrid model was trained by taking advantage of the Genetic Algorithm (GA). The results also indicated that the light condition has been the most significant parameter in evaluating the level of severity associated with fixed object crashes among elderly drivers, which is followed by the existence of the right and left shoulders. Following these three contributing factors, cause of collision, Average Annual Daily Traffic (AADT), number of involved vehicles, age, road surface condition, and gender have been identified as the most important variables in the developed ANN, respectively. This helps to identify gaps and improve public safety towards improving the overall highway safety situation of older drivers. [ABSTRACT FROM AUTHOR]

Energy use forecasting is crucial in balancing the electricity supply and demand to reduce the uncertainty inherent in the inter-basin water transfer project. Energy use prediction supports the reliable water-energy supply and encourages cost-effective operation by improving generation scheduling. The objectives are to develop subsequent monthly energy use predictive models for the Mokelumne River Aqueduct in California, US. Partial objectives are to (a) compare the model performance of a baseline model (multiple linear regression (MLR)) to three machine learning-based models (random forest (RF), deep neural network (DNN), support vector regression (SVR)), (b) compare the model performance of the whole system to three subsystems (conveyance, treatment, distribution), and (c) conduct sensitivity analysis. We simulate a total of 64 cases (4 algorithms (MLR, RF, DNN, SVR) x 4 systems (whole, conveyance, treatment, distribution) x 4 scenarios (different combinations of independent variables). We concluded that the three machine learning algorithms showed better model performance than the baseline model as they reflected non-linear energy use characteristics for water transfer systems. Among the three machine learning algorithms, DNN models yielded higher model performance than RF and SVR models. Subsystems performed better than the whole system as the models more closely reflected the unique energy use characteristics of the subsystems. The best case was having water supply (t), water supply (t-1), precipitation (t), temperature (t), and population (y) as independent variables. These models can help water and energy utility managers to understand energy performance better and enhance the energy efficiency of their water transfer systems. [ABSTRACT FROM AUTHOR]

Sedimentation is an important aspect of water resources management with many implications. Often, process‐based methods are employed to predict and assess the amount of sediment in water, but there are still challenges because the mechanisms that govern sediment transport are not yet fully understood. Furthermore, complex domains make model calibration difficult. Thus, as a complementary tool, a machine‐learning model was developed in the present study to emulate an existing process‐based model in simulating suspended sediment concentration (SSC). It employs the long short‐term memory (LSTM) networks, which are a type of artificial neural networks (ANNs) designed for supervised learning of a sequence of data (e.g., time series). The model was applied to the Sacramento‐San Joaquin Delta (the Delta) of California, USA, which is characterized by an interconnected system of sloughs, waterways, and a tidal outlet. The model training was performed with historical records of flow, stage and SSC at various locations within the Delta. The study period was 2010 through 2016, but the training period (i.e., range of observed data used to train the model) was varied to assess the model's sensitivity to the inputs and to determine the optimum model setup. Comparison between the model‐estimated SSC and the observation at 12 key locations within the Delta showed that the estimation accuracy of the LSTM model during the study period is comparable or superior to that of the Delta Simulation Model II‐General Transport Model (DSM2‐GTM), a process‐based operational hydrodynamics and water quality model for the Delta. In terms of the ratio of the root‐mean‐square error to the standard deviation (RSR), LSTM models generally showed higher predictability than DSM2‐GTM in all test cases investigated, with the lowest (most desirable) and highest (least desirable) LSTM‐based RSR being 0.21 and 1.14, respectively. In comparison, the lowest and highest RSR values with DSM2‐GTM were 0.26 and 3.70, respectively. The median LSTM‐based RSR of all study locations is around 0.7 while its DSM2‐GTM counterpart is about 1.0. LSTM models also yielded remarkably higher (more desirable) Nash‐Sutcliffe Efficiency values. Moreover, visual inspection found that LSTM models better captured the timing and magnitude of peaks as well as the temporal variations in the SSC time series. The LSTM model's performance was further analysed with hydro‐meteorological data (precipitation and wind speed) incorporated in training. While precipitation led to some improvement, the wind speed was found to have a negligible effect overall. All in all, the study findings suggest that the LSTM model has the potential to supplement the operational process‐based model in guiding water resources planning and management practices. [ABSTRACT FROM AUTHOR]

The levels of total salinity and its ion constituents in estuarine environments are important indicators of the overall suitability of water for environmental, agricultural, and urban use. These constituents include Total Dissolved Solids (TDS), dissolved chloride (Cl−), dissolved sulfate (SO42−), dissolved sodium (Na+), dissolved calcium (Ca2+), dissolved magnesium (Mg2+), dissolved nitrate (NO3), dissolved potassium (K), dissolved bromide (Br−), dissolved boron (B), Alkalinity, and hardness, among others. In practice, salinity is typically measured indirectly as electrical conductance (EC) via automatic sensors while the concentration of each constituent is often measured from discrete water samples (i.e., grab samples) and thus is available much less frequently than salinity. Quadratic regression equations are generally developed between salinity (as the predictor, represented by EC) and individual constituents (as the predictand) based on grab sample data. The regression models are then applied to estimate the concentrations of individual constituents given EC during the period when grab samples are not available. The current study develops four types of machine learning models: the Generalized Additive Model, Regression Trees, Random Forest, and Artificial Neural Networks, to emulate conventional regression equations in salinity-constituent conversion. A case study in the South Sacramento-San Joaquin Delta of California is presented to illustrate the development and application of these models and to compare their performance with that of the benchmark regression models. The results indicate that machine learning models can provide comparable or superior simulations to the regression models. Among the four machine models examined, the Random Forest model tends to yield the best results. The study further discusses the scientific and practical implications of the models proposed, their limitations, as well as future work to further improve their performance and applicability. [ABSTRACT FROM AUTHOR]

Accurate rainfall-runoff modelling is particularly challenging due to complex nonlinear relationships between various factors such as rainfall characteristics, soil properties, land use, and temporal lags. Recently, with improvements to computation systems and resources, data-driven models have shown good performances for runoff forecasting. However, the relative performance of common data-driven models using small temporal resolutions is still unclear. This study presents an application of data-driven models using artificial neural network, support vector regression and long-short term memory approaches and distributed forcing data for runoff predictions between 2010 and 2019 in the Russian River basin, California, USA. These models were used to predict hourly runoff with 1–6 h of lead time using precipitation, soil moisture, baseflow and land surface temperature datasets provided from the North American Land Data Assimilation System. The predicted results were evaluated in terms of seasonal and event-based performance using various statistical metrics. The results showed that the long-short term memory and support vector regression models outperforms artificial neural network model for hourly runoff forecasting, and the predictive performance of the models was greater during the wet seasons compared to the dry seasons. In addition, a comparison of the data-driven model results with the National Water Model, a fully distributed physical-based hydrologic model, showed that the long-short term memory and support vector regression models provide comparable performance. The results demonstrate that data-driven models for hourly runoff forecasting are sufficiently predictive and useful in areas where observation systems are not available. [ABSTRACT FROM AUTHOR]

Advances in machine-learning techniques can serve practical water management needs such as salinity level estimation. This study explores machine learning, particularly deep-learning techniques in developing computer emulators for a commonly used process model, the Delta Simulation Model II (DSM2), used for salinity estimation in California's Sacramento-San Joaquin Delta (Delta). We apply historical daily input data to DSM2 and corresponding salinity simulations at 28 study locations from 1990 to 2019 to train two machine-learning models: a multi-layer perceptron (MLP) and Long-Short-Term Memory (LSTM) networks in a multi-task learning framework. We assess sensitivity of both networks to the amount of antecedent input information (memory) and training data to determine appropriate memory size and training data length. We evaluate network performance according to several statistical metrics as well as visual inspection. The study further investigates two additional networks, the Gated Recurrent Unit (GRU) and Residual Network (ResNet) in salinity modeling, and compares their efficacy against MLP and LSTM. Our results demonstrate strong performance of the four neural network models over the study period, achieving absolute bias below 4%, plus near-perfect correlation coefficients and Nash–Sutcliffe efficiency coefficients. The high complexity LSTM shows slight performance edge. We further show that deeper and wider versions of MLP and LSTM yield only marginal benefit over their baseline counterparts. We also examined issues related to potential overfitting by the proposed models, training data selection strategies, and analytical and practical implications. Overall, this new study indicates that machine-learning-based emulators can efficiently emulate DSM2 in salinity simulation. They exhibit strong potential to supplement DSM2 in salinity modeling and help guide water resource planning and management practices for the Delta region. [ABSTRACT FROM AUTHOR]

In seismic structural health monitoring (SHM), a structure is normally instrumented with limited sensors at certain locations to monitor its structural behavior during an earthquake event. To reconstruct the responses at non-instrumented locations, an effective regression method has to be used given the measured data from the sensed locations. In addition, determination of where to place the sensors directly affects the ability of the system to infer the behaviour of the entire structure. In this study, a practical framework is proposed for sensor placement and seismic response reconstruction at non-instrumented locations, which adopts a novel attention-based deep neural network (DNN). The developed DNN model is trained by using structural displacements at measured locations as input and the structural displacements at unmeasured locations of interest as output. The proposed framework is demonstrated by a case study of an instrumented long-span girder bridge in California. Different sensor placement schemes are investigated using the proposed DNN model. Real-time seismic assessment of the bridge is achieved by issuing each reconstructed output in 1.5 ms. The case study validates the effectiveness and accuracy of the proposed method to be used as part of a seismic SHM system. [ABSTRACT FROM AUTHOR]

Summary: The ability of least square support vector machine (LSSVM) is investigated in this paper for modeling discharge-suspended sediment relationship. The daily stream flow and suspended sediment concentration data from two stations on the Eel River in California were used as case studies. In the first part of the study, the LSSVM was compared with those of the artificial neural networks (ANNs) and sediment rating curve (SRC) in the prediction of upstream and downstream station sediment data, separately. Two different algorithms, Levenberg–Marquardt and Conjugate Gradient, were employed for the ANN applications. For evaluating the ability of the models, root mean square errors, mean absolute errors and determination coefficient statistics were used. Comparison results showed that the LSSVM model was able to produce better results than the ANN models. LSSVM and ANN models were found to be better than the SRC model for the upstream station. For the downstream station, however, SRC model outperformed the LSSVM and ANN models. In the second part of the study, the models were compared to each other in estimation of downstream suspended sediment data by using data from both stations. It was found that the LSSVM model performs slightly better than the ANN models and both models performed much better than the SRC model. [ABSTRACT FROM AUTHOR]

Abstract: Traffic incidents inevitably cause traffic delay and deteriorate road safety conditions. Incidents are increasing alongside the fast economic growth. Due to the rampant growth of traffic incidents, developing efficient and effective automated incident detection (AID) techniques has prompted a growing worldwide interest. In this paper, the great efforts on developing a new approach to this problem based on nFOIL, a novel inductive logic programming (ILP), are done. By way of illustration, a simulated traffic data generated from Ayer Rajah Expressway (AYE) highway in Singapore and a real traffic data collected in I-880 freeway in California are used to assess the detection performance of this approach, and performance metrics includes detection rate, false alarm rate, mean time to detection, classification rate and the area under Receiver Operating Characteristic (ROC) curve (AUC). For comparison, we conducted the experiments on neural networks and support vector machine. The experimental results showed that nFOIL is sensitive to the skewed distribution of positive and negative examples in the dataset, and we make use of two different techniques, resampling and ensemble learning, to cope with highly skewed data in the context of ILP classification problems and investigated the effect of them typicality on the performance of AID model. It is concluded that ILP based AID approach are feasible, and have a favorable performance compared to neural networks and support vector machines. [Copyright &y& Elsevier]

Abstract: Estimating sediment volume carried by a river is an important issue in water resources engineering. This paper compares the accuracy of three different soft computing methods, Artificial Neural Networks (ANNs), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Gene Expression Programming (GEP), in estimating daily suspended sediment concentration on rivers by using hydro-meteorological data. The daily rainfall, streamflow and suspended sediment concentration data from Eel River near Dos Rios, at California, USA are used as a case study. The comparison results indicate that the GEP model performs better than the other models in daily suspended sediment concentration estimation for the particular data sets used in this study. Levenberg–Marquardt, conjugate gradient and gradient descent training algorithms were used for the ANN models. Out of three algorithms, the Conjugate gradient algorithm was found to be better than the others. [Copyright &y& Elsevier]

Abstract: Operating reserve (OR) is a major portion of ancillary services (AS) in a competitive electricity market and need to be procured by independent system operator (ISO), to achieve a high degree of power system reliability and security, following the major generation and transmission contingencies. Several ISOs have adopted deterministic methods to assess the OR requirements, however, such methods do not explicitly consider the unforeseen load swings and the probability of equipment outages. This paper proposes an adaptive wavelet neural network (AWNN) based two-stage approach to forecast OR requirements for both day-ahead and hour-ahead AS market in the California ISO (CAISO) controlled grid. The AWNN is a new class of feed-forward neural network with continuous wavelet function as the hidden layer node''s activation function. The forecasting results for winter and summer seasons of the year 2007 are presented and compared with those obtained by feed-forward multi-layer perceptron neural network (MLPNN). It is found that AWNN based proposed method outperforms the MLPNN model. [Copyright &y& Elsevier]

Summary: Correct estimation of sediment volume carried by a river is very important for many water resources projects. Conventional sediment rating curves, however, are not able to provide sufficiently accurate results. In this paper, an adaptive neuro-fuzzy approach is proposed to estimate suspended sediment concentration on rivers. The daily rainfall, streamflow and suspended sediment concentration data from Mad River Catchment near Arcata, USA are used as a case study. In the first part of the study, various combinations of current daily rainfall, streamflow and past daily streamflow, suspended sediment data are used as inputs to the neuro-fuzzy computing technique so as to estimate current suspended sediment. In the second part of the study, the potential of neuro-fuzzy technique is compared with those of the three different artificial neural networks (ANN) techniques, namely, the generalized regression neural networks (GRNN), radial basis neural networks (RBNN) and multi-layer perceptron (MLP) and two different sediment rating curves (SRC). The comparison results reveal that the neuro-fuzzy models perform better than the other models in daily suspended sediment concentration estimation for the particular data sets used in this study. [Copyright &y& Elsevier]

• A prediction methodology to predict wind and solar power curtailments in CAISO. • Machine learning-based forecasting using cross-validation and hold-out approaches. • Analyzing the six-year hourly data of power plants, power exchange, and demand. • A correlation matrix of eight input features and the three target variables. • Cross-validation improved the prediction of oversupply using random forest method. The economic viability of renewable energy is deteriorating due to its curtailment in power systems. Therefore, it is imperative to forecast curtailments for more effective utilization. To alleviate this issue, in this paper, we propose artificial intelligent-based models to accurately predict wind and solar power curtailments (WSPCs), which have not been investigated before. In this regard, a prediction methodology is developed using different types of machine learning (ML) methods and evaluated based on both hold-out (HO) and cross-validation (CV) approaches. The ML methods considered include regression trees (RT), gradient boosting trees (GBT), random forest (RF), feed-forward artificial neural networks (ANN), long short-term memory (LSTM), and support vector machines (SVR). The prediction models are trained based on eight input features, including load demands, the output power of thermal power plants, nuclear units, solar farms, wind turbines, biomass/geothermal units, large hydro units, power imports, and WSPC as two target variables. Based on historical data, i.e., hourly records of California independent system operator (ISO), the predictive models are validated, and the optimal hyperparameters are chosen using Bayesian optimization for each model to attain the best results. Among all the models, the RF model results in the minimum prediction errors and thus the best performance by implementing the proposed CV approach. The obtained results demonstrate the effectiveness of the proposed models in the prediction of WSPCs. [ABSTRACT FROM AUTHOR]

A one‐year's worth of near‐global marine shallow single‐layer cloud top radiative cooling (CTRC) is derived from a radiative transfer model with inputs from the satellite cloud retrievals and reanalysis sounding. The mean cloud top radiative flux divergence is 61 Wm−2, decomposed into the longwave and shortwave components of 73 and −11 W m−2, respectively. Equatorward of 30°N/S, the CTRC is largely a reflection of free‐atmospheric specific humidity distribution: a dry atmosphere enhances CTRC by reducing downward thermal radiation. Consequently, the cooling minimizes in the "wet" tropics and maximizes in the "dry" eastern subtropics. Poleward of 30°N/S, the CTRC decreases slightly due to the colder clouds that emit less effectively. The CTRC exhibits distinctive seasonal cycles with stronger cooling in the winter and has amplitudes of order 10–20 Wm−2 in stratocumulus‐rich regions. The datasets were used to train a machine‐learning model that substantially speeds up the retrieval. Plain Language Summary: Marine low‐lying clouds cool by emitting thermal radiation. The cooling is known as cloud top radiative cooling (CTRC). A change in CTRC can influence the properties of marine clouds via many avenues, ranging from altering the vertical motions of the clouds to changing the clouds' ability to reflect sunlight. Despite the importance of CTRC to the climate system, its climatological characteristics, namely how it varies with space and time, remain unknown. This work fills this knowledge gap. We generate the product of the CTRC over the global ocean using a novel satellite methodology developed in our previous work. Analyses of the data show that the spatial and temporal distributions of the CTRC are largely reflections of the atmospheric humidity: the drier the atmosphere, the stronger the cooling. As a result, the CTRC is weakest in the wet tropics and strongest in the dry eastern subtropical oceans, such as the west of California coast. We also use the CTRC data to train a machine‐learning algorithm that can substantially speed up the calculation of CTRC. Key Points: A one‐year's worth of global marine shallow single‐layer cloud top radiative cooling (CTRC) is derived from satellite and reanalysis dataSpatial and seasonal variations of CTRC are largely reflections of changes in free‐tropospheric humidity and cloud top temperatureA neural network model for the CTRC was trained, which substantially speeds up the retrieval while maintaining good accuracy [ABSTRACT FROM AUTHOR]

Domain-specific architectures of artificial neural networks (ANNs) have been developed to estimate salinity levels for planning at key monitoring stations in the Sacramento-San Joaquin Delta, California. In this work, we propose three major enhancements to existing ANN architectures for purposes of training time reduction, estimation error reduction, and better feature extraction. Specifically, we design a novel multitask ANN architecture with shared hidden layers for joint salinity estimation at multiple stations, achieving a reduction of 90% training and inference time. As another major structural redesign, we replace predetermined preprocessing on input data by a trainable convolution layer. We further enhance the multitask ANN design and training for salinity forecasting. Test results indicate that these enhancements substantially improve the efficiency and expand the capacity of the current salinity modeling ANNs in the Delta. Our enhanced ANN design methodologies have the potential for incorporation into the current modeling practice and provide more robust and timely information to guide water resource planning and management in the Delta. [ABSTRACT FROM AUTHOR]

Forecast skill of numerical weather prediction (NWP) models for precipitation accumulations over California is rather limited at subseasonal time scales, and the low signal-to-noise ratio makes it challenging to extract information that provides reliable probabilistic forecasts. A statistical post processing framework is proposed that uses an artificial neural network (ANN) to establish relationships between NWP ensemble forecast and gridded observed 7-day precipitation accumulations, and to model the increase or decrease of the probabilities for different precipitation categories relative to their climatological frequencies. Adding predictors with geographic information and location-specific normalization of forecast information permits the use of a single ANN for the entire forecast domain and thus reduces the risk of over fitting. In addition, a convolutional neural network (CNN) framework is proposed that extends the basic ANN and takes images of large-scale predictors as inputs that inform local increase or decrease of precipitation probabilities relative to climatology. Both methods are demonstrated with ECMWF ensemble reforecasts over California for lead times up to 4 weeks. They compare favorably with a state-of-the-art post processing technique developed for medium-range ensemble precipitation forecasts, and their forecast skill relative to climatology is positive everywhere within the domain. The magnitude of skill, however, is low for week-3 and week-4, and suggests that additional sources of predictability need to be explored. [ABSTRACT FROM AUTHOR]

Fundus images obtained in a telemedicine program are acquired at different sites that are captured by people who have varying levels of experience. These result in a relatively high percentage of images which are later marked as unreadable by graders. Unreadable images require a recapture which is time and cost intensive. An automated method that determines the image quality during acquisition is an effective alternative. To determine the image quality during acquisition, we describe here an automated method for the assessment of image quality in the context of diabetic retinopathy. The method explicitly applies machine learning techniques to access the image and to determine ‘accept’ and ‘reject’ categories. ‘Reject’ category image requires a recapture. A deep convolution neural network is trained to grade the images automatically. A large representative set of 7000 colour fundus images was used for the experiment which was obtained from the EyePACS that were made available by the California Healthcare Foundation. Three retinal image analysis experts were employed to categorise these images into ‘accept’ and ‘reject’ classes based on the precise definition of image quality in the context of DR. The network was trained using 3428 images. The method shows an accuracy of 100% to successfully categorise ‘accept’ and ‘reject’ images, which is about 2% higher than the traditional machine learning method. On a clinical trial, the proposed method shows 97% agreement with human grader. The method can be easily incorporated with the fundus image capturing system in the acquisition centre and can guide the photographer whether a recapture is necessary or not. [ABSTRACT FROM AUTHOR]

Long-term streamflow data are vital for analysis of hydrological droughts. Using an artificial neural network (ANN) model and nine tree-ring indices, this study reconstructed the annual streamflow of the Sacramento River for the period from 1560 to 1871. Using the reconstructed streamflow data, the copula method was used for bivariate drought analysis, deriving a hydrological drought return period plot for the Sacramento River basin. Results showed strong correlation among drought characteristics, and the drought with a 20-year return period (17.2 million acre-feet (MAF) per year) in the Sacramento River basin could be considered a critical level of drought for water shortages. [ABSTRACT FROM AUTHOR]

The study investigates the ability of artificial neural networks (ANN) with artificial bee colony (ABC) algorithm in daily reference evapotranspiration (ET) modeling. The daily climatic data, solar radiation, air temperature, relative humidity, and wind speed from two stations, Pomona and Santa Monica, in Los Angeles, USA, are used as inputs to the ANN-ABC model so as to estimate ET obtained using the FAO-56 Penman-Monteith (PM) equation. In the first part of the study, the accuracy of ANN-ABC models is compared with those of the ANN models trained with Levenberg-Marquardt (LM) and standard back-propagation (SBP) algorithms and those of the following empirical models: The California Irrigation Management System (CIMIS) Penman, Hargreaves, and Ritchie methods. The mean square error (MSE), mean absolute error (MAE) and determination coefficient (R) statistics are used for evaluating the accuracy of the models. Based on the comparison results, the ANN-ABC and ANN-LM models are found to be superior alternative to the ANN-SBP models. In the second part of the study, the potential of the ANN-ABC, ANN-LM, and ANN-SBP models in estimation ET using nearby station data is investigated. [ABSTRACT FROM AUTHOR]

This study investigates the ability of fuzzy genetic (FG) approach in modeling of reference evapotranspiration (ET0). The daily climatic data, solar radiation, air temperature, relative humidity, and wind speed from three stations, Windsor, Oakville and Santa Rosa, in central California, are used as inputs to the FG models to estimate ET0 obtained using the FAO-56 Penman-Monteith equation. A comparison is made between the estimates provided by the FG and those of the following empirical models: the California Irrigation Management System Penman, Hargreaves, Ritchie, and Turc methods. The FG results are also compared with the artificial neural networks. Root-mean-square errors (RMSE), mean-absolute errors (MAE), and correlation coefficient statistics are used as comparing criteria for the evaluation of the models’ performances. The comparison results reveal that the FG models are superior to the ANN and empirical models in modeling ET0 process. For the Windsor, Oakville, and Santa Rosa stations, it was found that the FG models with RMSEW=0.138, MAEW=0.098, and RW=0.999; RMSEO=0.144, MAEO=0.102, and RO=0.999; and RMSES=0.167, MAES=0.115, and RS=0.998 in test period is superior in modeling daily ET0 than the other models, respectively. [ABSTRACT FROM AUTHOR]

Two different artificial neural network (ANN) techniques, multi-layer perceptrons (MLP) and radial basis neural networks (RBNN), are employed in the estimation of monthly pan evaporation. The monthly climatic data, air temperature, solar radiation, wind speed, pressure and humidity, of three stations operated by the U.S. Environmental Protection Agency in California are used as inputs to the ANN models to estimate monthly evaporation. In the first part of the study, the accuracy of MLP and RBNN are compared with each other. The multi-linear regression (MLR) and the Stephens–Stewart (SS) methods are also considered for the comparison. The concern of second part of the study is to investigate the ability of neural computing techniques, MLR and SS models in evaporation estimation using data from nearby station(s). The performances of the models are evaluated using mean square errors, mean absolute relative errors and determination coefficient ( R2). The effect of periodicity on model’s estimation performance is also investigated. Based on the comparisons, it is found that the MLP and RBNN computing techniques could be employed successfully in modeling monthly evaporation process. [ABSTRACT FROM AUTHOR]

The present study was carried out to develop generalized artificial neural network (GANN) based reference crop evapotranspiration models corresponding to FAO-56 PM, FAO-24 Radiation, Turc, and FAO-24 Blaney–Criddle methods. The generalized ANN models were developed using the data from four California Irrigation Management Information System (CIMIS) stations, namely, Davis, Castroville, Mulberry, and West Side Field Station. The average weighted standard error of estimate (WSEE) for the developed models, namely, GANN (4-5-1), GANN (3-4-1), GANN (5-6-1), and GANN (6-7-1) corresponding to the FAO-24 Blaney–Criddle, FAO-24 Radiation, Turc, and FAO-56PM was 0.72, 0.85, 0.63, and 0.48 mm day-1, respectively. The developed ANN models were applied at 2 CIMIS stations namely, Lodhi and Fresno, without any local training. The average WSEE for models GANN (4-5-1), GANN (3-4-1), GANN (5-6-1), and GANN (6-7-1) was 0.68, 0.71, 0.65, and 0.46 mm day-1, respectively In addition, the GANN (4-5-1) model corresponding to FAO-24 Blaney–Criddle was directly applied to four Indian locations, namely, Hoshangabad, Gwalior, Jabalapur, and Pendra. The model gave the average WSEE of 0.57 mm day-1. Based on the results it was concluded that the GANN models can be used directly to predict evapotranspiration (ETo) under the arid conditions, since they performed better than the conventional evapotranspiration (ETo) estimation method. [ABSTRACT FROM AUTHOR]

The article presents a study on an intelligent simulator to examine why a price hike occurred in the California electricity crisis. A multiagent intelligent simulator (MAIS) is compared with other methods, such as neural network (NN) and genetic algorithm (GA), using a real data set on power trading related to the crisis. The comparison showed that MAIS and the other methods performed in predicting the price fluctuation of wholesale electricity. Findings indicate that the price hike was due to an increase in fuel prices and real demand.

In situ hyperspectral data obtained with a high spectral resolution radiometer were analysed for identification of six conifer species. Hyperspectral data were measured in the summer and late fall seasons at 15-20 cm above portions of tree canopies from both the sunlit and shaded sides. An artificial neural network algorithm was applied for identification purposes. Six types of transformation were applied to the hyperspectral reflectance data (R), preprocessed with a simple smoothing, followed by band aggregation. These include log(R), first derivative of R, first derivative of log(R), normalized R, first derivative of normalized R, and log(normalized R). First derivative of log(R) and first derivative of normalized R resulted in best species recognition accuracies with greater than 90% average accuracies, more than 20% greater than the average accuracy obtained from the pre-processed hyperspectral data. The effect of hyperspectral data taken from the shade sides of tree canopies can be minimized by applying normalization or by taking the derivatives after applying a logarithm to the pre-processed data. We found that a big difference in solar angle did not cause a noticeable difference in accuracies of species recognition. [ABSTRACT FROM AUTHOR]

Trained radial basis function networks are well-suited for use in extracting rules and explanations because they contain a set of locally tuned units. However, for rule extraction to be useful, these networks must first be pruned to eliminate unnecessary weights. The pruning algorithm cannot search the network exhaustively because of the computational effort involved. It is shown that using multiple pruning methods with smart ordering of the pruning candidates, the number of weights in a radial basis function network can be reduced to a small fraction of the original number. The complexity of the pruning algorithm is quadratic (instead of exponential) in the number of network weights. Pruning performance is shown using a variety of benchmark problems from the University of California, Irvine machine learning database. [ABSTRACT FROM AUTHOR]

Accurate information on crop distribution is of great importance for a range of applications including crop yield estimation, greenhouse gas emission measurement and management policy formulation. Fine spatial resolution (FSR) remotely sensed imagery provides new opportunities for crop mapping at a detailed level. However, crop classification from FSR imagery is known to be challenging due to the great intra-class variability and low inter-class disparity in the data. In this research, a novel hybrid method (OSVM-OCNN) was proposed for crop classification from FSR imagery, which combines a shallow-structured object-based support vector machine (OSVM) with a deep-structured object-based convolutional neural network (OCNN). Unlike pixel-wise classification methods, the OSVM-OCNN method operates on objects as the basic units of analysis and, thus, classifies remotely sensed images at the object level. The proposed OSVM-OCNN harvests the complementary characteristics of the two sub-models, the OSVM with effective extraction of low-level within-object features and the OCNN with capture and utilization of high-level between-object information. By using a rule-based fusion strategy based primarily on the OCNN's prediction probability, the two sub-models were fused in a concise and effective manner. We investigated the effectiveness of the proposed method over two test sites (i.e., S1 and S2) that have distinctive and heterogeneous patterns of different crops in the Sacramento Valley, California, using FSR Synthetic Aperture Radar (SAR) and FSR multispectral data, respectively. Experimental results illustrated that the new proposed OSVM-OCNN approach increased markedly the classification accuracy for most of crop types in S1 and all crop types in S2, and it consistently achieved the most accurate accuracy in comparison with its two object-based sub-models (OSVM and OCNN) as well as the pixel-wise SVM (PSVM) and CNN (PCNN) methods. Our findings, thus, suggest that the proposed method is as an effective and efficient approach to solve the challenging problem of crop classification using FSR imagery (including from different remotely sensed platforms). More importantly, the OSVM-OCNN method is readily generalisable to other landscape classes and, thus, should provide a general solution to solve the complex FSR image classification problem. [ABSTRACT FROM AUTHOR]