Search

Short-term electricity price forecasting in deregulated electricity markets has been studied extensively in recent years but without significant reduction in price forecasting errors. Also demand-side management and short-term scheduling operations in smart grids do not require strictly very accurate forecast and can be executed with certain practical price thresholds. This paper, therefore, presents a multikernel extreme learning machine (MKELM) for both short-term electricity price forecasting and classification according to some prespecified price thresholds. The kernel ELM does not require the hidden layer mapping function to be known and produces robust prediction and classification in comparison with the conventional ELM using random weights between the input and hidden layers. Further in the MKELM formulation, the linear combination of the weighted kernels is optimized using vaporization precipitation-based water cycle algorithm (WCA) to produce significantly accurate electricity price prediction and classification. The combination of MKELM and WCA is named as WCA-MKELM in this work. To validate the effectiveness of the proposed approach, three electricity markets, namely PJM, Ontario and New South Wales, are considered for electricity price forecasting and classification producing fairly accurate results. [ABSTRACT FROM AUTHOR]

Probabilistic forecasting is an appropriate tool that helps electricity markets participants to improve their decision making. Due to changes in electricity prices, the point forecasting accuracy cannot be guaranteed. Hence, participants are more interested in results of probabilistic forecasting methods such as prediction intervals method. In this paper, a hybrid approach for probabilistic electricity price forecasting is presented. This model is based on using improved clonal selection algorithm and extreme learning machine for neural networks training process and wavelet preprocess. The wavelet is utilized to decompose data into well behaved subsets, which increases accuracy of the model. Also, due to the high required computational time for training the neural networks, autocorrelation function is used to reduce the number of neural networks inputs. Finally, in order to evaluate the proposed probabilistic forecasting method, the Ontario and Australian electricity markets data are used. [ABSTRACT FROM AUTHOR]

This research paper presents a novel sequential wavelet-artificial neural network (ANN) with embedded ANN-particle swarm optimization (PSO) for short-term day-ahead forecasting of market clearing price (MCP) in the Indian energy exchange. A precise price forecasting helps suppliers to set up bidding strategies, make investment decisions and be cautious against risks. Conversely, consumers can use price forecasting to exploit appropriate power purchasing strategies for maximum utility utilization. Here the most influential historical data, namely purchase bid and MCP, are considered for training the feed-forward back-propagation neural network. The proposed model involves three sequential phases. Initially, the raw historical data are smoothened by removing the high-frequency components using a wavelet transform method which may enable better training of neural network. Then, ANN is used to train historical patterns. More number of trials is carried out, and the final weights that give the least training error are stored. In the final phase, the stored weights that are obtained from various trials are used as the initial population for the embedded ANN-PSO model. Here the performance of the proposed forecasting model is carried out using three error indices, namely mean absolute percentage error, normalized mean square error and error variance. [ABSTRACT FROM AUTHOR]

Price forecasting has become one of the main focuses of electric power market research efforts as price is the key index to evaluate the market competition efficiency and reflects the operation condition of electricity market decision making. The work presented in this paper makes use of local linear wavelet neural networks to find the market clearing price for a given period, which is based on similar days approach. The results obtained through simulation are compared to other evolutionary optimization techniques surfaced in the recent state-of-the-art literature, including wavelet neural network model. The results presented in this paper demonstrate the potential of the proposed approach and show its effectiveness for electricity price forecasting. [ABSTRACT FROM AUTHOR]

Electricity price is an important aspect in a restructured power market. Because of nonlinearity, nonstationary, and volatility of electricity price, it is highly essential to forecast the price on a short-term scale. This paper presents an efficient method based on a hybrid functional link dynamic neural network (DNN) trained by an adaptive robust unscented Kalman filter (UKF). The proposed method forecasts hourly prices for the day-ahead electricity market. The functional block helps to introduce nonlinearity by expanding the input space to higher-dimensional space through a basis function without using any hidden layer like multilayer perceptron structure. The DNN includes one or more infinite impulse response filters in the forward path providing feedback connections between outputs and inputs. This allows signal flow in both forward and backward directions, giving the network a dynamic memory useful to mimic dynamic systems. Also to improve the accuracy of the forecast, the noise covariance matrices of the UKF are adapted recursively. The proposed method is tested on PJM electricity market, and the residuals mean absolute error is compared with other forecasting methods, indicating the improved accuracy of the approach and its suitability to produce a real-time forecast. Further, to compare the accuracy of the forecast, an alternative UKF noise covariance optimization is attempted using differential evolution. [ABSTRACT FROM AUTHOR]

Machine learning neural network (NN) algorithms are being applied for the past few years in all engineering and science domain, economic sectors, image processing synthesis and analysis, and so on. Due to this, this paper work considered employing these machine learning neural models for forecasting application in respect of renewable energy applications and in particular focused on the forecasting of wind speed. It is crucial for power grid dispatchability, stability, and controllability, and its precision is necessary for making the most use of wind resources. This article describes the development of a novel hybrid forecasting system to anticipate the wind speed of real-time wind farm datasets using a hybrid probabilistic neural network (PNN) model and optimization method. The particle swarm optimization (PSO)–ant lion optimization technique is utilized to modify the designed adaptive PNN in order to optimize the weight parameters. The machine learning model employed in this study is an adaptive PNN with a probability density function and a decision-making function that adhere to Bayes' rule to achieve faster convergence and higher prediction accuracy. The obtained simulation results show that the recommended hybrid optimized PNN model outperforms the other techniques that were evaluated and compared from the literature. This establishes that the built optimized adaptive PNN model is applicable and suitable to serve as a predictor, as shown by the outcome of the statistical analysis. [ABSTRACT FROM AUTHOR]

As a common data-driven method, artificial neural networks have been widely used in electricity spot price forecasting. To improve the accuracy of short-term forecasts, this paper proposes an optimized artificial neural network model for monthly electricity spot prices forecasting. A genetic algorithm is applied to regulate the weights and biases parameters of the artificial neural network structure. This study uses various historical dataset at monthly periods selected from Nordic electricity spot prices. For efficiency comparison, one-step ahead forecast method based on Schwartz-Smith stochastic model and two other prediction models, artificial neural network trained by Levenberg–Marquardt and particle swarm optimization algorithms are also presented. The comparison results show that the prediction model based on the genetic optimization algorithm is more accurate than the other prediction models. The proposed forecasting model can be considered as an alternative technique for the electricity spot price forecasting in the Nordic regions. [ABSTRACT FROM AUTHOR]

In a smart grid, the main goals are to provide grid stability, improve power system performance and security, and reduce operations, system maintenance, and planning costs. The prediction stability of smart grid (SG) systems is essential in terms of power loss minimization and the importance of adequate energy policies. SG systems must accurately predict the energy demand and ensure the right amount of energy is available at the right time. If the prediction is inaccurate, it can lead to costly energy production or usage errors and create considerable inefficiencies in the power grid. Due to this, this manuscript offers five different cascade methods to detect the stability of SG systems. Detecting the stability of SG systems enables the grid to respond quickly to changes in demand and supply, improves system reliability, reduces power outages, and increases the overall efficiency of the grid. The present work proposed five different cascade methods with pre-processing, training and testing division, and the classification stages of the classification procedure for estimating SG stability. In the first pre-processing stage, the SG dataset is pre-proceeded with the feature selection (Relief, Fast Correlation-Based Filter (FCBF), and supervised attribute filter). The resampling (the bootstrapping), the Fuzzy C-Means Clustering-Based Feature Weighting (FCMFW), the resampling then feature selection (supervised attribute filter), and the feature selection (supervised attribute filter), then FCMFW. In the second stage, the training and testing division stage, the SG dataset was separated into three test and training data methods before the classification algorithm: The 5 Fold Cross Validation (FVC), 10 FVC, and hold-out (50–50%). In the third stage, the classification stage, five different classification algorithms, including Naive Kernel Bayes, Linear Support Vector Machine (SVM), Weighted K-Nearest Neighbors, Begged Trees, and Narrow Neural Network classifying algorithms, are used to classify the SG dataset. The simulation results of this study demonstrated that the suggested cascade ML system had achieved significant accuracy in predicting SG stability. The best cascade method is the feature selection (supervised attribute filter) + FCMFW + 10 FCV and then performing the bagged trees algorithm; thus, the new approach affords an accuracy of 99.9%. Furthermore, due to the rapid growth of ML techniques, sensors, and smart meters technologies, with Machine to Machine communication via the internet of things (IoT), the real-time identification process is made practical with higher accuracy. For this reason, our future research will focus on an IoT-based SG system, an E-stability determination system. Thanks to the proposed cascade method, the SG dataset can be classified easily, quickly, and reliably. E-stability determination systems can help to fast detect, predict, and respond, which is an important application of IoT on the grid systems. [ABSTRACT FROM AUTHOR]

Price forecasting of oil products and natural gas is of great interest due to their essential roles in modern industry and human lives. Existing studies on energy price forecasting are mostly concerned with individual energy markets, with less consideration of their interaction. This paper presents a bi-graph attention network (BiGAT) approach for energy price forecasting of oil products and natural gas, aiming at exploiting the price correlations of these energy sources for prediction. Specifically, we introduce the concordance graph and the causality graph into the BiGAT model to quantify the Kendall's rank correlation and the convergence cross mapping causality of the energy prices. To facilitate training the BiGAT model with the energy price time series of multiple time scale nature, we employ the boosted Hodrick–Prescott (bHP) filter to decompose the price data into slow- and fast-varying parts for learning, respectively. As the original bHP filtering algorithm involves computing an inverse matrix of the data size, it is not in favor of using in expanding or streaming data scenarios directly. Here, we also devise an incremental bHP filtering algorithm that applies to data of arbitrary finite size in a recursive manner, requiring the computation of a third-order initial inverse matrix only. Experimental results on empirical data show that the forecasting accuracy of our model is significantly better than other considered models, and the proposed incremental algorithm effectively extends the application scenarios of the bHP filter. [ABSTRACT FROM AUTHOR]

With the latest advances in deep learning-based generative models, it has not taken long to take advantage of their remarkable performance in the area of time series. Deep neural networks used to work with time series heavily depend on the size and consistency of the datasets used in training. These features are not usually abundant in the real world, where they are usually limited and often have constraints that must be guaranteed. Therefore, an effective way to increase the amount of data is by using data augmentation techniques, either by adding noise or permutations and by generating new synthetic data. This work systematically reviews the current state of the art in the area to provide an overview of all available algorithms and proposes a taxonomy of the most relevant research. The efficiency of the different variants will be evaluated as a central part of the process, as well as the different metrics to evaluate the performance and the main problems concerning each model will be analysed. The ultimate aim of this study is to provide a summary of the evolution and performance of areas that produce better results to guide future researchers in this field. [ABSTRACT FROM AUTHOR]

Tourist arrivals forecasting has become an increasingly hot issue due to its important role in the tourism industry and hence the whole economy of a country. However, owing to the complex characteristics of tourist arrivals series, such as seasonality, randomness, and non-linearity, forecasting tourist arrivals remains a challenging task. In this paper, a hybrid model of dual decomposition and an improved fuzzy time series method is proposed for tourist arrivals forecasting. In the novel model, two stages are mainly involved, i.e., dual decomposition and integrated forecasting. In the first stage, a dual decomposition strategy, which can overcome the potential defects of individual decomposition approaches, is designed to fully extract the main features of the tourist arrivals series and reduce the data complexity. In the second stage, a fuzzy time series method with fuzzy C-means algorithm as the discretization method is developed for prediction. In the empirical study, the proposed model is implemented to predict the monthly tourist arrivals to Hong Kong from USA, UK, and Germany. The results show that our hybrid model can obtain more accurate and more robust prediction results than benchmark models. Relative to the benchmark fuzzy time series models, the hybrid models using traditional decomposition methods and strategies, as well as the traditional single prediction models, our proposed model shows a significant improvement, with the improvement percentages at about 80, 70, and 50%, respectively. Therefore, we can conclude that the proposed model is a very promising tool for forecasting future tourist arrivals or other related fields with complex time series. [ABSTRACT FROM AUTHOR]

Knowing the demand for electrical consumption beforehand is important for efficient energy programming policies that can help with climate change, life cycle-costs, and optimal primary resource extraction. In this paper, we propose a framework to improve forecasting performance of high frequency electrical consumption data. We use different models for each day of the week, and then compose them to obtain the total forecast. We apply both machine learning (Long-Short Term Memory network) and econometric models (AutoRegressive Integrated Moving Average and Holtz-Winters) that consider time dependence in the data comparing model performance. We find that a classical ARIMA model outperforms other models; however, in applying the proposed framework, LSTM manages to outperform all other models. The results are statistically significant as indicated by the Model Confidence Set test constructed for Mean Absolute Percentage Error and Mean Square Error. [ABSTRACT FROM AUTHOR]

Agricultural price forecasting is one of the research hotspots in time series forecasting due to its unique characteristics. In this paper, we developed a deep long short-term memory (DLSTM) based model for the accurate forecasting of a nonstationary and nonlinear agricultural prices series. DLSTM model is a type of deep neural network which is advantageous in capturing the nonlinear and volatile patterns by utilizing both the recurrent architecture and deep learning methodologies together. The study further compares the price forecasting ability of the developed DLSTM model with conventional time-delay neural network (TDNN) and ARIMA models using international monthly price series of maize and palm oil. The empirical results demonstrate the superiority of the developed DLSTM model over other models in terms of various forecasting evaluation criteria like root mean square error, mean absolute percentage error and mean absolute deviation. The DLSTM model also showed dominance over other models in predicting the directional change of those monthly price series. Moreover, the accuracy of the forecasts obtained by all the models is also evaluated using the Diebold–Mariano test and the Friedman test whose results validate that the DLSTM based model has a clear advantage over the other two models. [ABSTRACT FROM AUTHOR]

A hybridized method that relies on using the support vector regression (SVR) method with equilibrium optimizer (EO) is proposed to foresee the closing prices of Egyptian Exchange (EGX). Three indices are modeled and employed: EGX 30, EGX 30 capped, and EGX 50 EWI. The efficiency of using the technical indicators and statistical measures in the forecasting process is evaluated. The proposed EO-SVR-based forecasting model is adopted and evaluated using mean absolute percentage error, average, standard deviation, best fit, worst fit, and CPU time. Also, it is compared with recently developed metaheuristic optimization algorithms published in the literature such as whale optimization algorithm, salp swarm algorithm, Harris Hawks optimization, gray wolf optimizer, Henry gas solubility optimization, Barnacles mating optimizer, Manta ray foraging optimization, and slime mold algorithm. The proposed EO-SVR model got better results than other the counterparts, and EO-SVR is considered the optimal model according to its superior outcomes. Moreover, there is no need to use technical indicators and statistical measures as their effect is not noticeable. [ABSTRACT FROM AUTHOR]

This work presents an analysis of Artificial Neural Network (ANN) based forecasting models for vehicular traffic flow. The forecasting was performed for three periods of time: 1 week, 1 month and 1 year. The vehicular traffic flow data analysed were collected from Interstate 87, a highway located within the U.S. state of New York. Temporal and climate information, such as weekday, time of day, precipitation, visibility and temperature, were provided to the ANNs. Two different architectures were implemented to forecast vehicular traffic flow: Multilayer Perceptron (MLP) and Long Short-Term Memory (LSTM). The performance was evaluated using the Root Mean Square Error (RMSE) and comparing the predicted flow with the real data in the evaluated period. The MLP resulted in RMSE of 113.96, 185.98 and 201.19, and MAP of 9.56%, 15.49% and 16.95% for a forecasting interval of 1 week, 1 month and 1 year, respectively. Similarly, the LSTM network resulted in RMSE of 122.20, 179.55 and 203.63 and MAPE of 14.66%, 17.89% and 24.10% for the same time periods. Additionally, the two developed ANNs were evaluated to perform a forecasting vehicular traffic flow for another highway, the Interstate 88. The results from the two ANNs are similar, the MLP network resulted in MAPE of 17.15%, 25.65% and 19.76% for a forecasting interval of 1 week, 1 month and 1 year, respectively. On the other hand, the LSTM network presented RMSE of 19.14%, 31.87% and 22.97%. The results obtained come close to the similar work results when considered shorter prediction periods. The results for longer prediction time could not be compared to the literature due to the lack of work in this area, being a novelty of the current work. Moreover, the models implemented in this paper were precise and stable and can be used as strategy information to help road and cities planning. In addition, results have shown that the model can be replicated for other roads in the same region or either from other states or countries. [ABSTRACT FROM AUTHOR]

In this paper, a novel combination of deep learning recurrent neural network and Lyapunov time is proposed to forecast the consumption of electricity load, in Greece, in normal/abrupt change value areas. Our method verifies the chaotic behavior of load time series through chaos time series analysis and with the application of deep learning recurrent neural networks produces predictions for 10 and 20 days ahead. Specifically, four different neural network models constructed (a) feed forward neural network, (b) gated recurrent unit (GRU) neural network, (c) long short-term memory (LSTM) recurrent and (d) bidirectional LSTM neural network to implement the prediction in a prediction horizon, produced through the extraction of maximum Lyapunov exponent. We constructed sequences of algorithms to feed the neural networks, creating three scenarios (a) 1-step, (b) 10-step and (c) 20-step sequences. For each neural network model, we used its predictions as inputs to predict steps forward, iteratively, to examine the accuracy of the proposed models, for horizons that are both inside and outside to that defined by Lyapunov time. The results show that the deep learning GRU neural network produces iterative predictions of high accuracy and stability, following the trend evolution of actual values, even outside the safe horizon for 1-step and 10-step cases. [ABSTRACT FROM AUTHOR]

Univariate time series forecasting is still an important but challenging task. Considering the wide application of temporal data, adaptive predictors are needed to study historical behavior and forecast future state in various scenarios. In this paper, inspired by human attention mechanism and decomposition and reconstruction framework, we proposed a convolution-based dual-stage attention (CDA) architecture combined with Long Short-Term Memory networks (LSTM) for univariate time series forecasting. Specifically, we first use the decomposition algorithm to generate derived variables from target series. Input variables are then fed into the CDA-LSTM machine for further forecasting. In the Encoder–Decoder phase, for the first stage, attention operation is combined with the LSTM acting as an encoder, which could adaptively learn the relevant derived series to the target. In the second stage, the temporal attention mechanism is integrated with decoder aiming to automatically select the relevant encoder hidden states across all time steps. A convolution phase is concatenated parallelly to the Encoder–Decoder phase to reuse the historical information of the target and extract the mutation features. The experimental results demonstrate the proposed method could be adopted as expert systems for forecasting in multiple scenarios, and the superiority is verified by comparing with twelve baseline models on ten datasets. The practicability of different decomposition algorithms and convolution architectures is also discussed by extensive experiment. Overall, our work carries a significant value not merely in adaptive modeling of deep learning in time series issues, but also in the field of univariate data processing and prediction. [ABSTRACT FROM AUTHOR]

In the process of building electrical load data collection, it is inevitable to introduce different kinds of noises, which makes the observation values deviate from the actual values, thus resulting in high levels of uncertainties. And such uncertainties make it difficult to achieve accurate point prediction of the short-term building electrical load. To improve the rationality of the prediction results and offer more effective information for decision makers, this paper proposes a novel multi-objective algorithm optimized modular fuzzy method which can accomplish the interval prediction for the short-term electrical load. First, one novel single-input-rule-modules (SIRMs)-based distributed interval fuzzy model (SIRM-DIFM) is proposed by replacing the original functional weights of the traditional SIRMs-based fuzzy inference system (SIRM-FIS) with the interval functional weights. Then, a data-driven learning scheme is presented for constructing the SIRM-DIFM. This learning sheme includes two main steps. The first step utilizes the iterative least square method to generate fuzzy rules for the SIRMs and determine the centers of the interval functional weights, while in the second step, the genetic algorithm (GA)-based multi-objective optimization algorithm is adopted to determine the widths of the interval functional weights. Through these two steps, accurate point estimation and reasonable interval prediction results can be achieved. Finally, two building electrical load prediction experiments are conducted to verify the effectiveness of the presented SIRM-DIFM. Simulation results indicate that the proposed SIRM-DIFM can compensate the shortcomings of the low accuracy of the point estimation and the predicted interval can effectively cover the observed data, providing the decision-makers more reliable and useful information. [ABSTRACT FROM AUTHOR]

Machine learning algorithms have been extensively exploited in energy research, due to their flexibility, automation and ability to handle big data. Among the most prominent machine learning algorithms are the boosting ones, which are known to be "garnering wisdom from a council of fools", thereby transforming weak learners to strong learners. Boosting algorithms are characterized by both high flexibility and high interpretability. The latter property is the result of recent developments by the statistical community. In this work, we provide understanding on the properties of boosting algorithms to facilitate a better exploitation of their strengths in energy research. In this respect, (a) we summarize recent advances on boosting algorithms, (b) we review relevant applications in energy research with those focusing on renewable energy (in particular those focusing on wind energy and solar energy) consisting a significant portion of the total ones, and (c) we describe how boosting algorithms are implemented and how their use is related to their properties. We show that boosting has been underexploited so far, while great advances in the energy field are possible both in terms of explanation and interpretation, and in terms of predictive performance. [ABSTRACT FROM AUTHOR]

Deep neural networks (DNNs) have evolved as a beneficial machine learning method that has been successfully used in various applications. Currently, DNN is a superior technique of extracting information from massive sets of data in a self-organized method. DNNs have different structures and parameters, which are usually produced for particular applications. Nevertheless, the training procedures of DNNs can be protracted depending on the given application and the size of the training set. Further, determining the most precise and practical structure of a deep learning method in a reasonable time is a possible problem related to this procedure. Meta-heuristics techniques, such as swarm intelligence (SI) and evolutionary computing (EC), represent optimization frames with specific theories and objective functions. These methods are adjustable and have been demonstrated their effectiveness in various applications; hence, they can optimize the DNNs models. This paper presents a comprehensive survey of the recent optimization methods (i.e., SI and EC) employed to enhance DNNs performance on various tasks. This paper also analyzes the importance of optimization methods in generating the optimal hyper-parameters and structures of DNNs in taking into consideration massive-scale data. Finally, several potential directions that still need improvements and open problems in evolutionary DNNs are identified. [ABSTRACT FROM AUTHOR]

Time series modelling is gaining spectacular popularity in the prediction process of decision making, with applications including real-world management and engineering. However, for short time series, prediction has to face unavoidable limitation for modelling extremely complex systems. It has to apply inadequate and incomplete data from short time to predict unknown observations. With such limited data source, existing techniques, such as statistical modelling or machine learning methods, fail to predict short time series effectively. To address this problem, this paper provides a global framework for short time series modelling predictions, integrating the rolling mechanism, grey model, and meta-heuristic optimization algorithms. In addition, dragonfly algorithm and whale optimization algorithm are investigated and deployed to optimize the framework by enhancing its predicting accuracy with less computational costs. To verify the performance of the proposed framework, three industrial cases are introduced as simulation experiments in this paper. Experimental results confirm that the framework solves corresponding short time series modelling predictions with greater accuracy and speed than the standard GM(1,1) models. The source codes of this framework are available at: https://github.com/zhesencui/HybridRollingGreyFramework.git. [ABSTRACT FROM AUTHOR]

Almost every organization around the globe is working with uncertainty due to inevitable changes and growth in every sphere of life. These changes affect directly or indirectly the stock market prices which makes forecasting a challenging task. So, the need for reliable, cost-effective, and accurate forecasting models significantly arises to reduce risk and uncertainty in stock market investment. Different time series models have been proposed by data scientists and researchers for accurate prediction of the future with the least errors. Econometric autoregressive time series models such as autoregressive moving average (ARMA) and autoregressive integrated moving average (ARIMA) models have established forecasting models capable of generating accurate forecasts. Wavelet methods, being capable of handling nonlinear data, combined with autoregressive models generate more accurate forecasts. In this present study, soft computing models of discreet wavelet transformation and wavelet denoising combined with autoregressive models are developed to forecast the weekly and daily closing prices of the BSE100 S&P Sensex index. Statistical error analysis of the forecasting outcomes of coupled models has been made to evaluate the performance of the prediction of these models. The prediction results reveal that soft computing methods coupled with autoregressive models (wavelet-ARIMA and wavelet denoise-ARIMA) generate considerably accurate forecasts as compared to baseline models (simple regression, ARMA and ARIMA models) and coupled models (wavelet-ARMA and wavelet denoise-ARMA models). [ABSTRACT FROM AUTHOR]

The analysis on economic data based on time series takes an important position in the field of analysis on time-series data and is also an important task of the field of big data and artificial intelligence. Traditional time-series analysis method is of relatively weak competence in dealing with multi-series analysis. In this research, based on the problem associated with the analysis on time-series economic data, efficient handling method and model are put forward in the face of multi-series analysis task. Also, combined with the association rules, trend correlation and self-trend correlation among multiple series, a trend and correlation deep neural network model (TC-DNM) is established and then tested and verified by using three kinds of economic datasets with representativeness based on the trend analysis task handed by multi-series analysis. The results show that the model proposed in this research is effective than a number of baseline models, can be employed to achieve precision–recall balance and also possesses strong reusability. The two correlation models and joint models in this paper are of peculiarity and innovativeness. [ABSTRACT FROM AUTHOR]

The purpose of smart city is to enhance the optimal utilization of scarce resources and improve the resident's quality of live. The smart cities employed Internet of Things (IoT) to create a sustainable urban life. The IoT devices such as sensors, actuators, and smartphones in the smart cities generate data. The data generated from the smart cities are subjected to analytics to gain insight and discover new knowledge for improving the efficiency and effectiveness of the smart cities. Recently, the application of deep learning in smart cities has gained a tremendous attention from the research community. However, despite raise in popularity and achievements made by deep learning in solving problems in smart cities, no survey has been dedicated mainly on the application of deep learning in smart cities to show recent progress and direction for future development. To bridge this gap, this paper proposes to conduct a dedicated survey on the applications of deep learning in smart cities. In this paper, recent progress, new taxonomies, challenges and opportunities for future research opportunities on the application of deep learning in smart cities have been unveiled. The paper can provide opportunities for experts in the research community to propose a novel approach for developing the research area. On the other hand, new researchers interested in the research area can use the paper as an entry point. [ABSTRACT FROM AUTHOR]

Electricity, a kind of clean energy, has been widely used in people's production and daily life. However, it is very difficult to estimate the electricity energy production in advance and store the rest of the electric energy due to the climate, environment, population and other factors. Based on data preprocessing and artificial intelligence optimization algorithm, this paper introduces a combined forecasting method. The proposed method contains six individual methods, and each individual method has its own usage. Singular spectrum analysis (SSA) is adopted to reduce noise from the original data; three individual forecasting methods, Jordan neural network, the echo state network, least squares support vector machine, are applied to obtain the intermediate forecasting results; two optimization algorithms, particle swarm optimization and simulated annealing, are used to optimize the parameters of the combined model. This paper not only validates the superiority of the combined model compared to the single predictive model through the simulation experiments of power load data and electricity price data. The mean absolute percent error (MAPE) of the combined power load and electricity price forecast results are 1.14% and 7.58%, respectively, which are higher than the MAPE error of the corresponding single models prediction results. It has also been verified that the process of eliminating noise by the SSA plays a positive role in the accuracy of the combined forecasting model. In addition, two series of experiments on the power load data lead to two very interesting conclusions. One of the conclusions is that as the size of the test data increases, the prediction accuracy of the model decreases; the other is that the predicted result calculated through the optimized combined weight is better than the combined result calculated using the average weight, and the average weight is used. Weighted combination does not improve the prediction accuracy of a single model. [ABSTRACT FROM AUTHOR]

Electricity demand forecasting plays a crucial role in the operation of electrical power systems because it can provide management decisions related to load switching and power grid. Thus, there have been models developed to estimate the electricity demand. However, inaccurate demand forecasting may raise the operating cost of electric power sector, which means that it would waste considerable money. In this paper, a novel modeling framework was proposed for forecasting electricity demand. Sample entropy was developed to identify the nonlinearity and uncertainty in the original time series, after that redundant noise was removed through a decomposition technique. Besides, the most optimal modes of original series and the optimal input form of the model were determined by the feature selection method. Finally, electricity demand series can be conducted forecasting through least squares support vector machine tuned by multi-objective sine cosine optimization algorithm. The case studies of Australia demonstrated that the proposed framework can ensure high accuracy and strong stability. Thus, it can be considered as a useful tool for electricity demand forecasting. [ABSTRACT FROM AUTHOR]

Stock price prediction is a rich research topic that has attracted interest from various areas of science. The recent success of machine learning in speech and image recognition has prompted researchers to apply these methods to asset price prediction. The majority of literature has been devoted to predicting either the actual asset price or the direction of price movement. In this paper, we study a hitherto little explored question of predicting significant changes in stock price based on previous changes using machine learning algorithms. We are particularly interested in the performance of neural network classifiers in the given context. To this end, we construct and test three neural network models including multilayer perceptron, convolutional net, and long short-term memory net. As benchmark models, we use random forest and relative strength index methods. The models are tested using 10-year daily stock price data of four major US public companies. Test results show that predicting significant changes in stock price can be accomplished with a high degree of accuracy. In particular, we obtain substantially better results than similar studies that forecast the direction of price change. [ABSTRACT FROM AUTHOR]

Forecasting in the field of power management is essential in recent days, due to the high electrical consumption at household and medical diagnosis applications to classify the electricity usage. It is highly impossible to identify the more accurate calculations in electricity consumption due to many uncertainties. This paper helps to overcome these uncertainties into probabilities by utilizing probabilistic neural network (PNN). The most complicated, complex and non-defined problems are well tackled by PNN as it is universally accepted as the best alternative technique. The conventional way of programming is not done but it is trained on the basis of behavioral representation of the data using the previous history. Multiple applications have been benefited using this system. Generally, PNN is used to differentiate four kinds of data produced from various grids and simultaneously the data of the grid are classified. 95% of reliability and accuracy is obtained from calculations produced from PNN as per the data results. The design can be used for appropriate grid development and to classify electricity usage. [ABSTRACT FROM AUTHOR]

A new hybrid model combining the kernel functions along with the random vector functional link neural network (RVFLN) is proposed in this paper for an effective solar power prediction. The conventional RVFLN is known for its fast learning speed, simple architecture and good generalization capabilities and allows direct connection between input and output nodes along with nonlinear enhancement nodes with random weights. However, the bottleneck of selecting the number of hidden enhancement nodes and mapping functions is still a challenging problem. To overcome these deficiencies of the conventional RVFLN, kernel functions are used for both the direct links and the hidden nodes to provide better stability, generalization and regression accuracy. Instead of using a single kernel for the enhancement nodes, this paper proposes an optimal kernel function that comprises a linear combination of weighted local kernel and a global kernel to improve the prediction accuracy of the solar power generation. This optimal kernel will be known as multi-kernel RVFLN (MK-RVFLN), and its parameters are optimized using an efficient metaheuristic evaporation-based water cycle (EVWCA-MKRVFLN) to provide accurate prediction of solar power. To validate its superior prediction accuracy, two solar power plants of 25 and 100 MW capacity in the states of New York and California are considered for 5-min- and 60-min-ahead prediction in the months of January, April, July and October. The result analysis shows that the MK-RVFLN algorithm attains better performance than many other techniques. [ABSTRACT FROM AUTHOR]

Many forecasting models are built on neural networks. The key issues in these models, which strongly translate into the accuracy of forecasts, are data representation and the decomposition of the forecasting problem. In this work, we consider both of these problems using short-term electricity load demand forecasting as an example. A load time series expresses both the trend and multiple seasonal cycles. To deal with multi-seasonality, we consider four methods of the problem decomposition. Depending on the decomposition degree, the problem is split into local subproblems which are modeled using neural networks. We move from the global model, which is competent for all forecasting tasks, through the local models competent for the subproblems, to the models built individually for each forecasting task. Additionally, we consider different ways of the input data encoding and analyze the impact of the data representation on the results. The forecasting models are examined on the real power system data from four European countries. Results indicate that the local approaches can significantly improve the accuracy of load forecasting, compared to the global approach. A greater degree of decomposition leads to the greater reduction in forecast errors. [ABSTRACT FROM AUTHOR]

This paper provides a new hybrid intelligent method for short-term prediction of the market-clearing price of electricity in the presence of wind power plants. The proposed method uses a data filtering technique based on wavelet transform and a radial basis function neural network, which is utilized for primary prediction. The main prediction engine comprises three MLP neural networks with different learning algorithms. To get rid of local minimums and to optimize the all neural networks, the meta-heuristic Imperialist Competitive Algorithm method is used. The input data for network training belong to the Nord Pool power market. The information includes a complete set of the historical record on electricity price and wind power generation. Moreover, the simultaneous impact of wind power generation is analyzed to predict the market-clearing price. Besides, the correlation coefficient factor is provided to consider the impact of wind power in forecasting the electricity price. Simulation results show the supremacy of the proposed method over other methods, to which it has been compared in this study. Also, the prediction error decreases significantly. [ABSTRACT FROM AUTHOR]

In this paper, we propose a novel feature extraction method for pattern classification problem. We propose to map the original data to subspaces for feature extraction and hope the mapped data can reconstruct the original data. The motive is to avoid of losing of information of the original data in the process of subspace mapping. We assume that if the original data can be reconstructed from the subspace, the critical information can be preserved. Moreover, we also observed that the reconstruction error is a low-rank matrix if the reconstruction is performed well. We propose to measure the reconstruction error matrix rank by the nuclear norm and minimize it to learn the optimal subspace transformation matrix. Meanwhile, the classification is also used to regularize the learning to improve the discriminate ability of the subspace representations. Experiments over several benchmark data sets show the advantage of the proposed method over the existing subspace learning methods. [ABSTRACT FROM AUTHOR]

This paper proposes a low-complexity pseudo-inverse Legendre neural network (PILNNR) with radial basis function (RBF) units in the hidden layer for accurate wind power prediction on a short-term basis varying from 10- to 60-min interval. The random input weights between the expanded input layer using Legendre polynomials and the RBF units in the hidden layer are optimized with a metaheuristic firefly (FF) algorithm for error minimization and improvement of the learning speed. For comparison, two other forecasting models, namely pseudo-inverse RBF (PIRBFNN-FF) neural network and PILNNR [with tanh functions in the hidden layer (PILNNT-FF)] with input-to-hidden layer weights being optimized by FF algorithm, are also presented in this paper. Also the weights between the hidden layer and the output neuron of these neural models are obtained by Moore–Penrose pseudo-inverse algorithm. Further to improve the stability of the weight learning procedure, the L2-norm-regularized least squares (ridge regression) technique is used. A superior predictive ability test is performed on the three proposed wind power forecasting models using bootstrapping procedure in order to identify the best model. Several case studies using wind power data of the wind farms in the states of Wyoming and California in USA and Sotavento wind farm in Spain are presented in this paper. [ABSTRACT FROM AUTHOR]

Over the years, high-dimensional, noisy, and time-varying natures of the stock markets are analyzed to carry out accurate prediction. Particularly, speculators and investors are understandably eager to accurately predict stock price since millions of dollars flow through the stock markets. At this point, soft computing models have empowered them to capture the data patterns and characteristics of stock markets. However, one of the open problems in soft computing models is how to systematically determine architecture of models for given applications. In this study, Harmony Search is utilized to optimize the architecture of Neural Network, Jordan Recurrent Neural Network, Extreme Learning Machine, Recurrent Extreme Learning Machine, Generalized Linear Model, Regression Tree, and Gaussian Process Regression for 1-, 2-, 3-, 5-, 7-, and 10-day-ahead stock price prediction. The experimental results show worthy findings of stock market behavior over different prediction terms and stocks. This study also helps researchers understand which prediction model performed the best and how different conditions affect the prediction accuracy of the models. Proposed hybrid models can be successfully used by speculators and investors to make the investment or to hedge against potential risk in stock markets. [ABSTRACT FROM AUTHOR]

Nearly one in five families in the country was poor in 2012, according to the Philippine Statistics Authority. While this proportion is lower than the corresponding figures from 2006 and 2009, the absolute number of poor families has actually grown from 3.8 million in 2006 to 4.2 million in 2012 due to the increase in population. Using data samples that have been collected from 69,130 households through a comprehensive community-based monitoring survey conducted in one of the cities that comprise Metro Manila, we attempt to identify the characteristics that differentiate between poor and non-poor households. Using back-propagation neural networks, we are able to correctly predict 73% of the poor households and 60% of the non-poor households. Moreover, the rules extracted from one of these networks provide concise description of how households are classified as poor based on their demographic characteristics and information pertaining to their surrounding living conditions. [ABSTRACT FROM AUTHOR]

A large number of renewable energies and uncertain power load accessing electric power system make the power load forecasting more complicated and face more new challenges. This paper presents a hybrid annual peak load forecasting model [namely MVO-DGM (1, 1)], which employs the latest optimization algorithm MVO (multi-verse optimizer) to determine two parameters of DGM (1, 1) model, and then uses the optimized DGM (1, 1) model to forecast annual peak load. The annual peak load of Shandong province in China from 2005 to 2014 is selected as the empirical example, and the analysis results demonstrate that the MVO algorithm for parameters’ determination of DGM (1, 1) model has significant superiority over the least square estimation method, particle swarm optimization and fruit fly optimization algorithm in terms of annual peak load forecasting. In addition, the proposed MVO-DGM (1, 1) peak load forecasting model has more excellent forecasting performance than other non-optimized forecasting techniques and other optimized DGM (1, 1) models due to its ascended local optima avoidance and better convergence speed. The hybrid MVO-DGM (1, 1) model proposed in this paper is feasible and effective in annual peak load forecasting, which can improve the forecasting accuracy. [ABSTRACT FROM AUTHOR]

This study compares time series and machine learning models for inflation forecasting. Empirical evidence from the USA between 1984 and 2014 suggests that out of sixteen conditions (four different inflation indicators and four different horizons), machine learning models provide more accurate forecasting results in seven conditions and the time series models are better in nine conditions. Moreover, multivariate models give better results in fourteen conditions, and univariate models are better only in two conditions. This study shows that machine learning model prevails against time series models for the core personal consumption expenditure (core-PCE) inflation forecasting, and the time series model (ARDL) is better for the core consumer price (core-CPI) index inflation forecasting in all horizons. [ABSTRACT FROM AUTHOR]

As a consequence of increasing wind power penetration level, it will be a big challenge to control and operate the power system because of the inherent uncertainty of the wind energy. One of the ways to deal with the wind power variability is to predict it accurately and reliably. The traditional point forecasting-based technique cannot notably solve the uncertainty in power system operation. In order to compute the probabilistic forecasting, which yields information on the uncertainty of wind power, a novel hybrid intelligent method that incorporates the wavelet transform, neural network (NN), and improved krill herd optimization algorithm (IKHOA), is used in this paper. Also, the extreme learning machine is exerted to train NN and calculates point forecasts, and IKHOA is applied to forecast the noise variance. The robust method called bootstrap is regarded to create prediction intervals and calculate the model uncertainty. The efficiency of proposed forecasting engine is evaluated by usage of wind power data from the Alberta, Canada. [ABSTRACT FROM AUTHOR]

In power industry and management, given the peculiarity and high complexity of the time series, it is highly requested to make models more flexible and well adapted to the data, in order to give them the ability to capture and recognize the most complex patterns, and consequently gain more accuracy in forecasting. To fulfill this aim, we define the seasonal autoregressive neural network (SAR-NN) as a dynamic feedforward artificial neural network (ANN), essentially conceived to forecast electricity prices. The SAR-NN is an ANN-based autoregressive model that considers that autoregressors are only those that are lagged by a multiple of the period p of the dominating seasonality. Moving forward this neural network, step-by-step, allows to generate multiple-steps-ahead reliable forecasts. This model is specifically designed to robustly overcome the strong seasonal effects and the other nonlinear patterns that often harm ANNs forecasting performance. The strategy is tested and compared to a number of homologous models throughout empirical experiments. As a case study, we focus on the Nord Pool Scandinavian power market, which is one of the most mature energy markets worldwide. [ABSTRACT FROM AUTHOR]

Multiplicative neuron model-based artificial neural networks are one of the artificial neural network types which have been proposed recently and have produced successful forecasting results. Sigmoid activation function was used in multiplicative neuron model-based artificial neural networks in the previous studies. Although artificial neural networks which involve the use of radial basis activation function produce more successful forecasting results, Gaussian activation function has not been used for multiplicative neuron model yet. In this study, rather than using a sigmoid activation function, Gaussian activation function was used in multiplicative neuron model artificial neural network. The weights of artificial neural network and parameters of activation functions were optimized by guaranteed convergence particle swarm optimization. Two major contributions of this study are as follows: the use of Gaussian activation function in multiplicative neuron model for the first time and the optimizing of central and propagation parameters of activation function with the weights of artificial neural network in a single optimization process. The superior forecasting performance of the proposed Gaussian activation function-based multiplicative neuron model artificial neural network was proved by applying it to real-life time series. [ABSTRACT FROM AUTHOR]

Forecasting electricity prices has been a widely investigated research issue in the deregulated power market scenario. High price volatilities, price spikes caused by a number of factors such as weather uncertainty, fluctuating fuel prices, transmission bottlenecks, etc., make the task of accurate price forecasting a formidable challenge for the market participants. A number of models have been proposed by researchers; however, achieving high accuracy is always not possible. In some specific applications such as self-scheduling by demand side participants, certain price thresholds are more useful than accurate price forecasts. In this paper, we have investigated the application of a novel neural network-based technique called extreme learning machine for the problem of classification of future electricity prices with respect to certain price thresholds. Different models corresponding to different lead times are developed and tested with data corresponding to Ontario and PJM markets. It is observed that classification with ELM is fast, less sensitive to user defined parameters and easily implementable. [ABSTRACT FROM AUTHOR]

In this paper, a novel self-adaptive extreme learning machine (ELM) based on affinity propagation (AP) is proposed to optimize the radial basis function neural network (RBFNN). As is well known, the parameters of original ELM which developed by G.-B. Huang are randomly determined. However, that cannot objectively obtain a set of optimal parameters of RBFNN trained by ELM algorithm for different realistic datasets. The AP algorithm can automatically produce a set of clustering centers for the different datasets. According to the results of AP, we can, respectively, get the cluster number and the radius value of each cluster. In that case, the above cluster number and radius value can be used to initialize the number and widths of hidden layer neurons in RBFNN and that is also the parameters of coefficient matrix H of ELM. This may successfully avoid the subjectivity prior knowledge and randomness of training RBFNN. Experimental results show that the method proposed in this thesis has a more powerful generalization capability than conventional ELM for an RBFNN. [ABSTRACT FROM AUTHOR]

Stock index forecasting is one of the most difficult tasks that financial organizations, firms and private investors have to face. Support vector regression (SVR) has become a popular alternative in stock index forecasting tasks due to its generalization capability in obtaining a unique solution. However, the major limitation of SVR is that it cannot capture the relative importance of independent variables to the dependent variable when many potential independent variables are considered. This study incorporates feature selection method and SVR for building stock index forecasting model. The proposed model uses multivariate adaptive regression splines (MARS), an effective nonlinear and nonparametric regression methodology, to identify important forecasting variables. The obtained significant predictor variables are then served as the inputs for the SVR model. Experimental results reveal that the obtained important variables from MARS can improve the forecasting performance of the SVR models. Moreover, the MARS results provide useful information about the relationship between the selected predictor variables and stock index through the obtained basis functions, important predictor variables and the MARS prediction function. Hence, the proposed stock index forecasting model can generate good forecasting performance and exhibits the capability of identifying significant predictor variables, which provide valuable information for further investment decisions/strategies. [ABSTRACT FROM AUTHOR]

The increased deregulation of electricity markets in most nations of the world in recent years has made it imperative that electricity utilities design accurate and efficient mechanisms for determining locational marginal price (LMP) in power systems. This paper presents a comparison of two soft computing-based schemes: Artificial neural networks and support vector machines for the projection of LMP. Our system has useful power system parameters as inputs and the LMP as output. Experimental results obtained suggest that although both methods give highly accurate results, support vector machines slightly outperform artificial neural networks and do so with manageable computational time costs. [ABSTRACT FROM AUTHOR]

As a novel and promising learning technology, extreme learning machine (ELM) is featured by its much faster training speed and better generalization performance over traditional learning techniques. ELM has found applications in solving many real-world engineering problems, including those in electric power systems. Maintaining frequency stability is one of the essential requirements for secure and reliable operations of a power system. Conventionally, its assessment involves solving a large set of nonlinear differential-algebraic equations, which is very time-consuming and can be only carried out off-line. This paper firstly reviews the ELM's applications in power engineering and then develops an ELM-based predictor for real-time frequency stability assessment (FSA) of power systems. The inputs of the predictor are power system operational parameters, and the output is the frequency stability margin that measures the stability degree of the power system subject to a contingency. By off-line training with a frequency stability database, the predictor can be online applied for real-time FSA. Benefiting from the very fast speed of ELM, the predictor can be online updated for enhanced robustness and reliability. The developed predictor is examined on the New England 10-generator 39-bus test system, and the simulation results show that it can exactly (within acceptable errors) and rapidly (within very small computing time) predict the frequency stability. [ABSTRACT FROM AUTHOR]

Electricity spot prices are complex processes characterized by nonlinearity and extreme volatility. Previous work on nonlinear modeling of electricity spot prices has shown encouraging results, and we build on this area by proposing an Expectation Maximization algorithm for maximum likelihood estimation of recurrent neural networks utilizing the Kalman filter and smoother. This involves inference of both parameters and hyper-parameters of the model which takes into account the model uncertainty and noise in the data. The Expectation Maximization algorithm uses a forward filtering and backward smoothing (Expectation) step, followed by a hyper-parameter estimation (Maximization) step. The model is validated across two data sets of different power exchanges. It is found that after learning a posteriori hyper-parameters, the proposed algorithm outperforms the real-time recurrent learning and the extended Kalman Filtering algorithm for recurrent networks, as well as other contemporary models that have been previously applied to the modeling of electricity spot prices. [ABSTRACT FROM AUTHOR]