Search

Despite its great importance, there has been no general consensus on how to model the trends in time-series data. Compared to traditional approaches, neural networks (NNs) have shown some promise in time-series forecasting. This paper investigates how to best model trend time series using NNs. Four different strategies (raw data, raw data with time index, detrending, and dif- ferencing) are used to model various trend patterns (linear, non- linear, deterministic, stochastic, and breaking trend). We find that with NNs differencing often gives meritorious results regardless of the underlying data generating processes (DGPs). This finding is also confirmed by the real gross national product (GNP) series. [ABSTRACT FROM AUTHOR]

The objective of this paper is to present a new methodology for midterm energy forecasting. The proposed model is an adaptive artificial neural network (ANN), which properly transforms the input variables to differences or relative differences, in order to predict energy values not included in the training set. The ANN parameters, such as the finally used input variables, the number of neurons, initial values, and time periods of momentum term and training rate, are simultaneously selected by an optimization process. Another characteristic of the model is the use of a minimal training set of patterns. Results from an extensive analysis conducted by the developed method for the Greek power system and for different categories of customers are compared to those obtained from the application of standard regression methods. [ABSTRACT FROM AUTHOR]

Short-term load modeling and forecasting are essential for operating power utilities profitably and securely. Modern machine learning approaches, such as neural networks, have been used for this purpose. This paper proposes using the alternative technique of abductive networks, which offers the advantages of simplified and more automated model synthesis and analytical input-output models that automatically select influential inputs, provide better insight and explanations, and allow comparison with statistical and empirical models. Using hourly temperature and load data for five years, 24 dedicated models for forecasting next-day hourly loads have been developed. Evaluated on data for the sixth year, the models give an overall mean absolute percentage error (MAPE) of 2.67%. Next-hour models utilizing available load data up to the forecasting hour give a MAPE of 1.14%, outperforming neural network models for the same utility data. Two methods of accounting for the load growth trend achieve comparable performance. Effects of varying model complexity are investigated and proposals made for further improving forecasting performance. [ABSTRACT FROM AUTHOR]

Accurately predicting the state of health (SoH) of lithium-ion batteries remains one of the top challenges, especially in uncertain environments where feature availability is not guaranteed. The main contribution of this article is to propose a highly available feature fault-tolerant prediction framework capable of predicting the SoH of batteries with a very high accuracy under sensor data unavailability. Previously published studies mitigated the feature incompleteness issue using models that deeply depend on a specific set of features; if one of the features becomes unavailable, the prediction model fails. Unlike the existing methods, the proposed feature fault-tolerant prediction framework proposes a novel agile approach so that when one of the features is missing, the model considers the next available feature to make the prediction. Our fault-tolerant prediction framework is highly available for SoH prediction and always predicts at a very high accuracy level. A hybrid deep learning model combining long short-term memory and deep neural networks along with a meticulous feature engineering design is used to train the NASA dataset and predict the SoH of batteries with a high precision. In order to prove the efficiency of the model, the feature fault-tolerant prediction framework configuration has been validated against two other datasets: Toyota Research Institute in partnership with MIT and Stanford dataset and NASA dynamic profile datasets. The performance is evaluated using the error indicators: mean absolute error, root-mean-square error, and the $R$ square. [ABSTRACT FROM AUTHOR]

In recent years, a large variety of short-term load forecasting methodologies have been proposed, but a common drawback of them is that require a large amount of historical data to train the model. However, in reality, the cases in limited historical data would be more frequently occurred, for example, a new house or substation is built. To tackle this issue, a combining forecasting method of deep residual neural network (ResNet)-based transfer learning for residential loads is proposed in this article. To improve the accuracy and reliability of transfer learning, a novel deep ResNet with dual skip connections (ResNet-DSC) is proposed as the base model. Then with sparse training data, a Bayesian probability weighted averaging (approach is proposed, to address the model combination parameters estimation problem. In addition, probability density function forecasting is also delivered by the method. For demonstration, a series of experiments, with actual residential load data, was performed. With the comparison with other transfer learning and nontransfer learning approaches, the effectiveness and improvement of the proposed method have been validated. [ABSTRACT FROM AUTHOR]

This letter proposes a transferrable model-agnostic meta-learning (T-MAML) approach for short-term load forecasting for single households. The proposed approach enables multiple households to collaboratively train a generic artificial neural network (ANN) model. The generic ANN model is then further trained at each target household node for the STLF purpose. The proposed T-MAML based STLF approach is featured by: (1) significant reduction of computation and communication costs on the household side; and (2) superior STLF performance, especially when there is limited load data for training in a target household. Experiments based on a real Australian residential dataset are conducted to validate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Predictive business process monitoring aims at forecasting potential problems during process execution before they occur so that these problems can be handled proactively. Several predictive monitoring techniques have been proposed in the past. However, so far those prediction techniques have been assessed only independently from each other, making it hard to reliably compare their applicability and accuracy. We empirically analyze and compare three main classes of predictive monitoring techniques, which are based on machine learning, constraint satisfaction, and Quality-of-Service (QoS) aggregation. Based on empirical evidence from an industrial case study in the area of transport and logistics, we assess those techniques with respect to five accuracy indicators. We further determine the dependency of accuracy on the point in time during process execution when a prediction is made in order to determine lead-times for accurate predictions. Our evidence suggests that, given a lead-time of half of the process duration, all predictive monitoring techniques consistently provide an accuracy of at least 70%. Yet, it also becomes evident that the techniques differ in terms of how accurately they may predict violations and nonviolations. To improve the prediction process, we thus exploit the characteristics of the individual techniques and propose their combination. Based on our case study data, evidence indicates that certain combinations of techniques may outperform individual techniques with respect to specific accuracy indicators. Combining constraint satisfaction with QoS aggregation, for instance, improves precision by 14%; combining machine learning with constraint satisfaction shows an improvement in recall by 23%. [ABSTRACT FROM AUTHOR]

The work in this article uses artificial neural networks and support vector machine to forecast electricity consumption in an industrial facility. The main objective is to show that such a problem should be treated with a contextual approach that identifies the most adequate technic in each moment for a single building, contrary to the previous works in the literature that compare the accuracy of each method for the complete data set representing aggregated loads. 72 different algorithms have been implemented and tested. After that, the three most suitable ones are selected in order to support the automated decisions of the best algorithm according to the context. In this way, the implemented methodology finds the best method for the prediction of each 5 min. It can be later used to update the production planning in the industrial facility. It also discussed the size of historical data and the most suitable learning parameters for each method. The case study includes test data for one week and more than one year of training data. [ABSTRACT FROM AUTHOR]

Implementation of a commercial application to a grid infrastructure introduces new challenges in managing the quality-of-service (QoS) requirements, most stem from the fact that negotiation on QoS between the user and the service provider should strictly be satisfied. An interesting commercial application with a wide impact on a variety of fields, which can benefit from the computational grid technologies, is three-dimensional (3-D) rendering. In order to implement, however, 3-D rendering to a grid infrastructure one should develop appropriate scheduling and resource allocation mechanisms so that the negotiated (QoS) requirements are met.

Electricity price is a crucial element for market players to maximize their profits. In this context, the forecast of the hour-ahead, day-ahead, and week-ahead electricity prices plays a crucial role. The more accurate the prediction is, the lower the market risk is. In this article, several machine learning algorithms (support vector machine, Gaussian processes regression, regression trees, and multilayer perceptron) with different structures have been adopted to forecast Italian wholesale electricity prices. Considering different time horizons (hourly, daily, and weekly), their performances have been compared through several performance metrics, including mean absolute error, R-index, mean absolute percentage error, and the number of anomalies in which the forecast error passes a threshold. The investigation reveals that, in general, SVM and tree-based models outperform other models at different time horizons. [ABSTRACT FROM AUTHOR]

A motion cueing algorithm (MCA) is employed to transform the linear and angular motion signals generated from a motion simulator without violating the physical and dynamical boundaries of the motion platform. In this respect, the accurate prediction of the motion scenarios is essential to enhance the efficiency of the MCA using prepositioning or time-varying reference model predictive control. While a recent approach that utilizes a feedforward neural network (NN) to forecast the motion scenarios is useful, the feedforward NN model has only forward dynamics relating to the signals without any feedback loop. In this article, a time-delay feedforward NN, a recurrent NN, and a nonlinear autoregressive (NAR) models with three different training procedures, i.e., Levenberg–Marquardt, Bayesian regularization, and scaled conjugate gradient, are exploited to predict the motion scenarios. As the NAR model employs the historical signals as the inputs, it can predict the motion scenarios with higher accuracy rates. Based on the series of empirical evaluations, NAR trained with Levenberg–Marquardt is able to outperform the other two counterparts in producing more accurate predictions of the motion signals. The NAR method has a lower computational load as compared with that of the recurrent NN, facilitating its real-time application. In addition to the MCA, the NAR method can be employed in other areas, including autonomous vehicles and motion sickness studies. It can also be easily implemented for air, sea, and/or land vehicle simulators for training purposes in virtual reality environments. [ABSTRACT FROM AUTHOR]

Presents information on a study which proposed a neural-network-based method of demand forecasting. Approaches for customer demand modeling; Description of neural-network-based demand model; Computational example.

Uncertainty modeling of renewable energy sources, load demand, electricity price, etc. create a high volume of data in smart grids. Accordingly, in this article, a precise forecasting method based on a deep learning concept with microclustering (MC) task is presented. The MC method is structured based on hybrid unsupervised and supervised clustering tasks by K-means and Gaussian support vector machine, respectively. In the proposed method, the input data sequence is clustered by the MC task, and then the forecasting process is employed. By applying the MC, input data in each hour are categorized into different groups, and a distinctive forecasting unit is allocated to each one. In this way, more clusters and forecasting networks are earmarked for the hours with higher fluctuation rates. The bi-directional long short-term memory (B-LSTM), which is one of the newest recurrent artificial neural networks, is proposed as the forecasting unit. The B-LSTM has bidirectional memory—feedforward and feedback loops—that helps us to investigate both previous and future hidden layers data. The optimal number of clusters in each hour is determined based on the Davies–Bouldin index. To evaluate the performance of the proposed method, in this study, three forecasting tasks including the wind speed, load demand, and electricity price are studied in different periods using the Ontario province, Canada, data set. The results are compared with other benchmarking methods to verify the robustness and effectiveness of the proposed method. In fact, the proposed method, which is equipped with the MC technique and B-LSTM networks, significantly promotes the forecasting results, especially in spike points. [ABSTRACT FROM AUTHOR]

The article presents and summarizes the research results of the UTC-UTC(PL) differences predicting procedure for the Polish Time Scale UTC(PL) developed at the Institute of Metrology, Electronics and Computer Science of the University of Zielona Góra. The research has been conducted in close cooperation with the Time and Frequency Laboratory of Central Office of Measures, responsible for the realization of the UTC(PL). The planned experiment covered a 50-month period, in which the UTC(PL) timescale has been carried out using, as a master clock, atomic clocks with different properties: commercial cesium atomic clock 5071A Opt. 001 (first case) and VCH-1003M hydrogen maser (second and third cases). In case of the developed predicting procedure by means of group method of data handling (GMDH)-type neural network (NN), the transitional period (second case) in which the master clock had been replaced by the new clock, is also very important. For GMDH-type NN, input data as a TS1 and TS2 time series have been prepared, based on the already-known differences of UTC(PL) from UTC and UTC rapid timescales and on determined differences between UTC(PL) and the master clock steering the UTC(PL). The method of preparing TS1 and TS2 time series is an important part of the developed procedure for predicting UTC(PL) differences from UTC, before they will be calculated and published by fr. Bureau International des Poids et Mesures (BIPM). The obtained research results confirmed that the developed predicting procedure by means of GMDH-type NN ensures a proper quality of maintaining the national timescale. The best quality of UTC(PL) predicting has been achieved for the third case, both for data prepared according to the TS1 and TS2 time series. This is confirmed by the calculated values of the prediction quality measures and the Modified Allan deviation. On the other hand, the second case confirmed the possibility of using the proprietary development of the TS2 time series to predict UTC-UTC(k) differences. This time series is a very good alternative for the TS1 time series, when the amount of data in this series is insufficient due to replacement of the clock or its failure. [ABSTRACT FROM AUTHOR]

Presents a study which described a method for input variable selection for artificial neural network (ANN) based short-term electric load forecasting (STLF). Different methods for identification of ANN; Phase-space embedding procedure; Phase-space reconstruction results.

This article proposes an optimal energy scheduling method for power transmission networks using novel genetic algorithm (nGA) for solving the dynamic economic dispatch (DED) problem combined with machine learning based short-term load forecasting (STLF). The STLF is implemented based on a multilayer artificial neural network (MANN) to estimate the day-ahead variations in the demand. The efficacy of the proposed energy scheduling model together with the STLF is verified using a modified IEEE 30-bus system using real data of the power plants located in the Ereymentau region of Kazakhstan. The simulation results suggest that the proposed model offers a cost effective, reliable, and efficient dynamic energy scheduling in power transmission systems. [ABSTRACT FROM AUTHOR]

Reports on the use of cascaded artificial neural networks (CANN) for short-term electric load forecasting. Discussion of artificial neural networks (ANN); ANN configuration; Case study for the examining the effectiveness of the proposed CANN approach; Conclusions.

Investigates the ability to solve the short-term load forecasting (STLF) problem with Artificial Neural Networks (ANN) by conducting a fractional factorial experiment. Testing results and analysis; Rules for building a `quasi optimal network'; Comparison of quasi optimal neural networks versus an automated Box-Jenkins seasonal ARIMA model to solve thr STLF problem; Conclusions.

Focuses on electric Short Term Load Forecasting (STLF) based on Artificial Neural Network (ANN) architectures. STF of the Italian electric hourly load; Day-type classification by means of unsupervised learning; Formulation of the learning algorithm; Description of test activity.

The predispatch price forecast plays a key element in the electricity market. However, such a forecast usually depends on the traditional offline batch-learning technologies, which cannot respond in time to the unexpected changes in the local power system environment. Further, the predispatch local price forecast is often affected by the dynamic price changes from the neighboring regions. This article proposes a novel online learning forecast approach to overcome the above issues to provide a better predispatch price forecast by using the online sequential extreme learning machine (OS-ELM) algorithm. The article proposes a novel data structure in the form of a 2-D orthogonal list and two corresponding OS-ELM modules. One module provides the rolling day-ahead price prediction and prediction intervals using the day-by-day online training update, while the other provides the rolling 30-min prediction using the 2-h-by-2-h online training update. The proposed approach can continuously perceive any unexpected events and any price fluctuations from the neighboring regions in the nonlinear patterns. The proposed approach is validated using simulation studies based on the data from the Australian electricity market, and the simulation results show that the proposed approach can help in improving the forecast accuracy, especially when unexpected changes occur both locally and in the neighboring area. [ABSTRACT FROM AUTHOR]

While automatic controllers are frequently used during transit operations and low-speed maneuvering of ships, ship operators typically perform docking maneuvers. This task is more or less challenging depending on factors, such as local environment disturbances, the number of nearby vessels, and the speed of the ship as it docks. This article proposes a tool for onboard support that offers position predictions based on an integration of a supervised machine learning (ML) model of the ship into the ship dynamic model. The ML model is applied as a compensator of the unmodeled behavior or inaccuracies from the dynamic model. The dynamic model increases the amount of predetermined knowledge about how the vessel is likely to move and, thus, reduces the black-box factor typically experienced in purely data-driven predictors. A prediction horizon of 30 s ahead of real time during docking operations is examined. History data from the 29-m coastal displacement ship Research Vessel (RV) Gunnerus are applied to validate the approach. Results show that the inclusion of the data-based ML model significantly improves the prediction accuracy. [ABSTRACT FROM AUTHOR]

A machine learning-based application for estimating a reflected electromagnetic spectrum is presented in this article. Since the results of the spectrum estimation by our previously implemented cubic Hermite interpolation method had some drawbacks, and in line with the increasing application of machine learning in sensor technology, the new hypothesis is put forward, based on the assumption that better predictions could be achieved by introducing artificial neural networks. The input data of such networks consist of six measured values, which represent the intensities of reflected light at specific wavelengths, while the output is formed out of 36 points, which are used to predict the shape of the spectral curve in the range of 380–730 nm (to obtain a resolution of 10 nm). Colorimetric capabilities of the proposed method are compared with a commercial spectrophotometer and with our previously obtained results using the cubic Hermite interpolation. The machine learning-based method has proven to be an intelligent solution for the spectrum estimation, and the drawbacks from the previously used method were reduced significantly. [ABSTRACT FROM AUTHOR]

As the population keeps growing, traffic congestion happens more and more often. Consequently, travel time has become an important indicator of driving experience. Accurate travel time information helps drivers plan their route more wisely and thus effectively alleviate traffic congestion. In this research, we propose a vehicle travel time prediction model for freeway traffic. The data used in this research are derived from the traffic dataset of the Taiwan Freeway Bureau, and the travel time prediction is made for the Sun Yat-sen Freeway between Taipei and Hsinchu. First, the missing value of the raw data is imputed by Autoencoder. The data are then segmented according to time series and are used to build the prediction model. To effectively capture the hidden features required to predict the travel time for the vehicle traveling on the freeway, a deep learning architecture is adopted in our system, which includes the GRU neural network model, the XGBoost model, and the Hybrid model that combines the GRU and XGBoost through linear regression. To increase computational efficiency, the travel time predictions for consecutive toll gates every 5 minutes apart are pre-computed offline, so that the online travel time prediction of the whole trip can be obtained by simply summing up a few numbers. Experimental results based on actual traffic data show that the proposed system can achieve good performance in terms of prediction accuracy and execution time. [ABSTRACT FROM AUTHOR]

With the increase of solar photovoltaic (PV) penetration in power system, the impact of random fluctuation of PV power on the secure operation of power grid becomes more and more serious. High-precision PV power forecasting can effectively promote the grid's accommodation of PV power generation. Cloud is the most important factor affecting the surface irradiance and PV power. For the ultra-short-term solar PV power forecast considering the influence of cloud movement, it is necessary to be able to obtain the surface irradiance according to the sky cloud observation data. Therefore, in order to accurately achieve the real-time mapping relationship between sky image and surface irradiance, a hybrid mapping model based on deep learning applied for solar PV power forecasting is proposed in this article. First, the sky image data are clustered based on the feature extraction of convolutional autoencoder and K-means clustering algorithm after preprocess stage. Second, a hybrid mapping model based on deep learning methods are established for surface irradiance. Finally, the simulation results are compared and evaluated with different deep learning methods (CNN, LSTM, and ANN). The results show that the proposed model in this article has higher accuracy and can maintain robustness under different weather conditions. [ABSTRACT FROM AUTHOR]

This letter proposes a novel granule computing-based framework for prediction intervals (PIs) construction of solar irradiance time series that has significant impacts on solar power production. Distinguished from most existing methods, the new framework can address both stochastic and knowledge uncertainties in constructing PIs. The proposed method has proved to be highly effective in terms of both reliability and sharpness through a real case study using measurement data obtained from Hong Kong Observatory. [ABSTRACT FROM PUBLISHER]

A short-term forecasting of the electricity price with data-driven algorithms is studied in this research. A stacked denoising autoencoder (SDA) model, a class of deep neural networks, and its extended version are utilized to forecast the electricity price hourly. Data collected in Nebraska, Arkansas, Louisiana, Texas, and Indiana hubs in U.S. are utilized. Two types of forecasting, the online hourly forecasting and day-ahead hourly forecasting, are examined. In online forecasting, SDA models are compared with data-driven approaches including the classical neural networks, support vector machine, multivariate adaptive regression splines, and least absolute shrinkage and selection operator. In the day-ahead forecasting, the effectiveness of SDA models is further validated through comparing with industrial results and a recently reported method. Computational results demonstrate that SDA models are capable to accurately forecast electricity prices and the extended SDA model further improves the forecasting performance. [ABSTRACT FROM AUTHOR]

A novel methodology for probabilistic wind power forecasting is described. The method is based on artificial intelligence and concentrates on the uncertainty information about the future wind power production predicting a set of quantiles with predefined nominal probabilities. The proposed model uses the point predictions of an existing state-of-the-art wind power forecasting model and forecasts the prediction uncertainties due to the inaccuracies of the numerical weather predictions (NWP), the weather stability and the deterministic forecasting model. The performance of the proposed model is evaluated on two wind farms that are located in areas with different weather conditions. [ABSTRACT FROM PUBLISHER]

Time series has been a popular tool for the analysis and forecasting of a large number of data. Very often, the applied approaches forecasts had limited success and the main reason was the lack of statistically significant historical information. We focus our attention on three common series, which are formed from the averaging of data collected over a shorter time interval. These include weekly and biweekly foreign exchange rates, mean hourly wind speed and electric load data. The proposed scheme, which takes advantage of the dominant characteristics of the shorter interval data, produced superior forecasts to those based on conventional approaches based only on historical observations of the target data. In the first two series, the proposed approach generated forecasts that significantly lower to those of the trivial random walk, a benchmark in series dominated by short-term correlation. On the load series, this approach made possible that a simple Auto-Regressive model returned lower forecasting error compared to a neural network that included special indicators to account for the periodic nature of the data. [ABSTRACT FROM AUTHOR]

Average load forecasting errors for the holidays are much higher than those for weekdays. So far, many studies on the short-term load forecasting have been made to improve the prediction accuracy using various methods such as deterministic, stochastic, artificial neural net (ANN) and neural network-fuzzy methods. In order to reduce the load forecasting error of the 24 hourly loads for the holidays, the concept of fuzzy regression analysis is employed in the short-term load forecasting problem. According to the historical load data, the same type of holiday showed a similar trend of load profile as in previous years. The fuzzy linear regression model is made from the load data of the previous three years and the coefficients of the model are found by solving the mixed linear programming problem. The proposed algorithm shows good accuracy, and the average maximum percentage error is 3.57% in the load forecasting of the holidays for the years of 1996-1997. [ABSTRACT FROM AUTHOR]

A polynomial neural network (PNN) is first applied to construct a predictive model of plasma etch processes. The process was characterized by a full-factorial experiment, and two attributes modeled are an etch rate and a dc bias. The training and test data for the etch rate consisted of 32 and 15 patterns, respectively. For the dc bias, the training and test data were composed of 34 and 17 patterns, respectively. Prediction performance of PNN was optimized with a variation in the number of input factors and in the polynomial type. For each data type, 27 cases were evaluated. The root mean squared prediction accuracy is 8.49 nm/min and 5.43 V for the optimized etch rate and dc-bias models, respectively. For comparison, backpropagation neural network (BPNN) and three types of statistical regression models were constructed. Five training factors involved in training the BPNN were experimentally optimized. Compared to other models, the PNN model demonstrated an improvement of more than 30% and 80% in modeling the etch rate and dc bias, respectively. By the demonstrated high prediction ability, the PNN can be effectively used to model and control complex plasma processes. [ABSTRACT FROM AUTHOR]

Presents a study which focused on the combination of artificial neural-network forecasters with application to the prediction of daily natural gas consumption needed by gas utilities. Methods of forecast combination; Results; Conclusions.

Presents information on a study designed to apply neural-networks to the problem of forecasting the flow of the River Nile in Egypt. Utilization of the time series as a benchmark; Comparison of four different forecasting methods to preprocess the inputs and outputs; Comparison of three different methods for multistep ahead forecast problem.

Presents a study which used artificial neural networks (NN's) to estimate rainfall from remotely sensed data. General background about NN's; Description of the data used in the study; Examination of the performance of the NN and regression approaches in estimating rainfall; Comparison of these results to radar-estimated rainrates; Suggestions for future work given.

Presents the development of an artificial neural network based short-term load forecasting model for an autonomous power system. Basic artificial neural network structure; Input data processing of short-term load forecasting in the island of Crete, an autonomous power system; Sensitivity of model to temperature; Discussion of test results; Conclusions.

Rapidly evolving businesses generate massive amounts of time-stamped data sequences and cause a demand for both univariate and multivariate time series forecasting. For such data, traditional predictive models based on autoregression are often not sufficient to capture complex nonlinear relationships between multidimensional features and the time series outputs. In order to exploit these relationships for improved time series forecasting while also better dealing with a wider variety of prediction scenarios, a forecasting system requires a flexible and generic architecture to accommodate and tune various individual predictors as well as combination methods. In reply to this challenge, an architecture for combined, multilevel time series prediction is proposed, which is suitable for many different universal regressors and combination methods. The key strength of this architecture is its ability to build a diversified ensemble of individual predictors that form an input to a multilevel selection and fusion process before the final optimized output is obtained. Excellent generalization ability is achieved due to the highly boosted complementarity of individual models further enforced through cross-validation-linked training on exclusive data subsets and ensemble output postprocessing. In a sample configuration with basic neural network predictors and a mean combiner, the proposed system has been evaluated in different scenarios and showed a clear prediction performance gain. [ABSTRACT FROM AUTHOR]