Abstract: Forecasting electricity load is important for various purposes like planning, operation and control. Forecasts can save operating and maintenance costs, increase the reliability of power supply and delivery systems, and correct decisions for future development. This paper compares various time series methods to forecast 24 hours ahead of electricity load. The methods considered are the Holt-Winters smoothing, SARIMA Modeling, LSTM Network, Fbprophet and Tensorflow probability. The performance of each method is evaluated by using the forecasting accuracy criteria namely, the Mean Absolute Error and Root Mean Square Error. The National Renewable Energy Laboratory (NREL) residential energy consumption data are used to train the models. The results of this study show that SARIMA model is superior to the others for 24 hours ahead forecasts. Furthermore, a Bagging technique is used to make the predictions more robust. The obtained results show that by Bagging multiple time-series forecasts we can improve the robustness of the models for 24 hour ahead electricity load forecasting.
Abstract: Time Series Forecasting (TSF) is used to predict the target variables at a future time point based on the learning from previous time points. To keep the problem tractable, learning methods use data from a fixed length window in the past as an explicit input. In this paper, we study how the performance of predictive models change as a function of different look-back window sizes and different amounts of time to predict into the future. We also consider the performance of the recent attention-based transformer models, which had good success in the image processing and natural language processing domains. In all, we compare four different deep learning methods (Recurrent Neural Network (RNN), Long Short-term Memory (LSTM), Gated Recurrent Units (GRU), and Transformer) along with a baseline method. The dataset (hourly) we used is the Beijing Air Quality Dataset from the website of University of California, Irvine (UCI), which includes a multivariate time series of many factors measured on an hourly basis for a period of 5 years (2010-14). For each model, we also report on the relationship between the performance and the look-back window sizes and the number of predicted time points into the future. Our experiments suggest that Transformer models have the best performance with the lowest Mean Absolute Errors (MAE = 14.599, 23.273) and Root Mean Square Errors (RSME = 23.573, 38.131) for most of our single-step and multi-steps predictions. The best size for the look-back window to predict 1 hour into the future appears to be one day, while 2 or 4 days perform the best to predict 3 hours into the future.
Abstract: This study explores the practicality of using electroencephalographic (EEG) independent components to predict eight-direction finger movements in pseudo-real-time. Six healthy participants with individual-head MRI images performed finger movements in eight directions with two different arm configurations. The analysis was performed in two stages. The first stage consisted of using independent component analysis (ICA) to separate the signals representing brain activity from non-brain activity signals and to obtain the unmixing matrix. The resulting independent components (ICs) were checked, and those reflecting brain-activity were selected. Finally, the time series of the selected ICs were used to predict eight finger-movement directions using Sparse Logistic Regression (SLR). The second stage consisted of using the previously obtained unmixing matrix, the selected ICs, and the model obtained by applying SLR to classify a different EEG dataset. This method was applied to two different settings, namely the single-participant level and the group-level. For the single-participant level, the EEG dataset used in the first stage and the EEG dataset used in the second stage originated from the same participant. For the group-level, the EEG datasets used in the first stage were constructed by temporally concatenating each combination without repetition of the EEG datasets of five participants out of six, whereas the EEG dataset used in the second stage originated from the remaining participants. The average test classification results across datasets (mean ± S.D.) were 38.62 ± 8.36% for the single-participant, which was significantly higher than the chance level (12.50 ± 0.01%), and 27.26 ± 4.39% for the group-level which was also significantly higher than the chance level (12.49% ± 0.01%). The classification accuracy within [–45°, 45°] of the true direction is 70.03 ± 8.14% for single-participant and 62.63 ± 6.07% for group-level which may be promising for some real-life applications. Clustering and contribution analyses further revealed the brain regions involved in finger movement and the temporal aspect of their contribution to the classification. These results showed the possibility of using the ICA-based method in combination with other methods to build a real-time system to control prostheses.
Abstract: Amman, the capital of Jordan, is the main political, economic, social and cultural center of Jordan and beyond. The city faces multitude demographic challenges related to the unstable political situation in the surrounded countries. It has regional and local migrants who left their homes to find better life in the capital. This resulted with random and unequaled population distribution. Some districts have high population and pressure on the infrastructure and services more than other districts.Government works to resolve this challenge in compliance with 100 Cities Resilience Framework (CRF). Amman participated in this framework as a member in December 2014 to work in achieving the four goals: health and welfare, infrastructure and utilities, economy and education as well as administration and government. Previous research studies lack in studying Amman resilient work in neighborhood scale and the population growth as resilient challenge. For that, this study focuses on Al-Thahabiah neighborhood in Shafa Badran district in Amman. This paper studies the reasons and drivers behind this population growth during the selected period in this area then provide strategies to improve the resilient work in neighborhood scale. The methodology comprises of primary and secondary data. The primary data consist of interviews with chief officer in the executive part in Great Amman Municipality and resilient officer. The secondary data consist of papers, journals, newspaper, articles and book’s reading. The other part of data consists of maps and statistical data which describe the infrastructural and social situation in the neighborhood and district level during the studying period. Based upon those data, more detailed information will be found, e.g., the centralizing position of population and the provided infrastructure for them. This will help to provide these services and infrastructure to other neighborhoods and enhance population distribution. This study develops an analytical framework to assess urban demographical time series in accordance with the criteria of CRF to make accurate detailed projections on the requirements for the future development in the neighborhood scale and organize the human requirements for affordable quality housing, employment, transportation, health and education in this neighborhood to improve the social relations between its inhabitants and the community. This study highlights on the localization of resilient work in neighborhood scale and spread the resilient knowledge related to the shortage of its research in Jordan. Studying the resilient work from population growth challenge perspective helps improve the facilities provide to the inhabitants and improve their quality of life.
Abstract: In fields such as neuroscience and especially in cognition modeling of mental processes, uncertainty processing in temporal zone of signal is vital. In this paper, Bayesian online inferences in estimation of change-points location in signal are constructed. This method separated the observed signal into independent series and studies the change and variation of the regime of data locally with related statistical characteristics. We give conditions on simulations of the method when the data characteristics of signals vary, and provide empirical evidence to show the performance of method. It is verified that correlation between series around the change point location and its characteristics such as Signal to Noise Ratios and mean value of signal has important factor on fluctuating in finding proper location of change point. And one of the main contributions of this study is related to representing of these influences of signal statistical characteristics for finding abrupt variation in signal. There are two different structures for simulations which in first case one abrupt change in temporal section of signal is considered with variable position and secondly multiple variations are considered. Finally, influence of statistical characteristic for changing the location of change point is explained in details in simulation results with different artificial signals.
Abstract: Generative Adversarial Net (GAN) has proved to be a powerful machine learning tool in image data analysis and generation. In this paper, we propose to use Conditional Generative Adversarial Net (CGAN) to learn and simulate time series data. The conditions include both categorical and continuous variables with different auxiliary information. Our simulation studies show that CGAN has the capability to learn different types of normal and heavy-tailed distributions, as well as dependent structures of different time series. It also has the capability to generate conditional predictive distributions consistent with training data distributions. We also provide an in-depth discussion on the rationale behind GAN and the neural networks as hierarchical splines to establish a clear connection with existing statistical methods of distribution generation. In practice, CGAN has a wide range of applications in market risk and counterparty risk analysis: it can be applied to learn historical data and generate scenarios for the calculation of Value-at-Risk (VaR) and Expected Shortfall (ES), and it can also predict the movement of the market risk factors. We present a real data analysis including a backtesting to demonstrate that CGAN can outperform Historical Simulation (HS), a popular method in market risk analysis to calculate VaR. CGAN can also be applied in economic time series modeling and forecasting. In this regard, we have included an example of hypothetical shock analysis for economic models and the generation of potential CCAR scenarios by CGAN at the end of the paper.
Abstract: Forecasting the stock market is a very challenging task. Various economic indicators such as GDP, exchange rates, interest rates, and unemployment have a substantial impact on the stock market. Time series models are the traditional methods used to predict stock market changes. In this paper, a machine learning method, Bayesian Additive Regression Tree (BART) is used in predicting stock market indexes based on multiple economic indicators. BART can be used to model heterogeneous treatment effects, and thereby works well when models are misspecified. It also has the capability to handle non-linear main effects and multi-way interactions without much input from financial analysts. In this research, BART is proposed to provide a reliable prediction on day-to-day stock market activities. By comparing the analysis results from BART and with time series method, BART can perform well and has better prediction capability than the traditional methods.
Abstract: Waste reduction is a fundamental problem for sustainability. Methods for waste reduction with point-of-sales (POS) data are proposed, utilizing the knowledge of a recent econophysics study on a statistical property of POS data. Concretely, the non-stationary time series analysis method based on the Particle Filter is developed, which considers abnormal fluctuation scaling known as Taylor's law. This method is extended for handling incomplete sales data because of stock-outs by introducing maximum likelihood estimation for censored data. The way for optimal stock determination with pricing the cost of waste reduction is also proposed. This study focuses on the examination of the methods for large sales numbers where Taylor's law is obvious. Numerical analysis using aggregated POS data shows the effectiveness of the methods to reduce food waste maintaining a high profit for large sales numbers. Moreover, the way of pricing the cost of waste reduction reveals that a small profit loss realizes substantial waste reduction, especially in the case that the proportionality constant of Taylor’s law is small. Specifically, around 1% profit loss realizes half disposal at =0.12, which is the actual value of processed food items used in this research. The methods provide practical and effective solutions for waste reduction keeping a high profit, especially with large sales numbers.
Abstract: Compared with terrestrial network, the traffic of spatial information network has both self-similarity and short correlation characteristics. By studying its traffic prediction method, the resource utilization of spatial information network can be improved, and the method can provide an important basis for traffic planning of a spatial information network. In this paper, considering the accuracy and complexity of the algorithm, the spatial information network traffic is decomposed into approximate component with long correlation and detail component with short correlation, and a time series hybrid prediction model based on wavelet decomposition is proposed to predict the spatial network traffic. Firstly, the original traffic data are decomposed to approximate components and detail components by using wavelet decomposition algorithm. According to the autocorrelation and partial correlation smearing and truncation characteristics of each component, the corresponding model (AR/MA/ARMA) of each detail component can be directly established, while the type of approximate component modeling can be established by ARIMA model after smoothing. Finally, the prediction results of the multiple models are fitted to obtain the prediction results of the original data. The method not only considers the self-similarity of a spatial information network, but also takes into account the short correlation caused by network burst information, which is verified by using the measured data of a certain back bone network released by the MAWI working group in 2018. Compared with the typical time series model, the predicted data of hybrid model is closer to the real traffic data and has a smaller relative root means square error, which is more suitable for a spatial information network.
Abstract: The statistical modelling of precipitation data for a
given portion of territory is fundamental for the monitoring of
climatic conditions and for Hydrogeological Management Plans
(HMP). This modelling is rendered particularly complex by the
changes taking place in the frequency and intensity of precipitation,
presumably to be attributed to the global climate change. This paper
applies the Wakeby distribution (with 5 parameters) as a theoretical
reference model. The number and the quality of the parameters
indicate that this distribution may be the appropriate choice for
the interpolations of the hydrological variables and, moreover, the
Wakeby is particularly suitable for describing phenomena producing
heavy tails. The proposed estimation methods for determining the
value of the Wakeby parameters are the same as those used for
density functions with heavy tails. The commonly used procedure
is the classic method of moments weighed with probabilities
(probability weighted moments, PWM) although this has often shown
difficulty of convergence, or rather, convergence to a configuration
of inappropriate parameters. In this paper, we analyze the problem of
the likelihood estimation of a random variable expressed through its
quantile function. The method of maximum likelihood, in this case,
is more demanding than in the situations of more usual estimation.
The reasons for this lie, in the sampling and asymptotic properties of
the estimators of maximum likelihood which improve the estimates
obtained with indications of their variability and, therefore, their
accuracy and reliability. These features are highly appreciated in
contexts where poor decisions, attributable to an inefficient or
incomplete information base, can cause serious damages.
Abstract: As a great physiographic divide, the Himalayas affecting a large system of water and air circulation which helps to determine the climatic condition in the Indian subcontinent to the south and mid-Asian highlands to the north. It creates obstacles by defending chill continental air from north side into India in winter and also defends rain-bearing southwesterly monsoon to give up maximum precipitation in that area in monsoon season. Nowadays extreme weather conditions such as heavy precipitation, cloudburst, flash flood, landslide and extreme avalanches are the regular happening incidents in the region of North Western Himalayan (NWH). The present study has been planned to investigate the suitable model(s) to find out the rainfall pattern over that region. For this investigation, selected models from Coordinated Regional Climate Downscaling Experiment (CORDEX) and Coupled Model Intercomparison Project Phase 5 (CMIP5) has been utilized in a consistent framework for the period of 1976 to 2000 (historical). The ability of these driving models from CORDEX domain and CMIP5 has been examined according to their capability of the spatial distribution as well as time series plot of rainfall over NWH in the rainy season and compared with the ground-based Indian Meteorological Department (IMD) gridded rainfall data set. It is noted from the analysis that the models like MIROC5 and MPI-ESM-LR from the both CORDEX and CMIP5 provide the best spatial distribution of rainfall over NWH region. But the driving models from CORDEX underestimates the daily rainfall amount as compared to CMIP5 driving models as it is unable to capture daily rainfall data properly when it has been plotted for time series (TS) individually for the state of Uttarakhand (UK) and Himachal Pradesh (HP). So finally it can be said that the driving models from CMIP5 are better than CORDEX domain models to investigate the rainfall pattern over NWH region.
Abstract: The Greek Energy Market is structured as a mandatory pool where the producers make their bid offers in day-ahead basis. The System Operator solves an optimization routine aiming at the minimization of the cost of produced electricity. The solution of the optimization problem leads to the calculation of the System Marginal Price (SMP). Accurate forecasts of the SMP can lead to increased profits and more efficient portfolio management from the producer`s perspective. Aim of this study is to provide a comparative analysis of various machine learning models such as artificial neural networks and neuro-fuzzy models for the prediction of the SMP of the Greek market. Machine learning algorithms are favored in predictions problems since they can capture and simulate the volatilities of complex time series.
Abstract: Back propagation algorithm (BP) is a widely used
technique in artificial neural network and has been used as a tool
for solving the time series problems, such as decreasing training
time, maximizing the ability to fall into local minima, and optimizing
sensitivity of the initial weights and bias. This paper proposes an
improvement of a BP technique which is called IM-COH algorithm
(IM-COH). By combining IM-COH algorithm with cuckoo search
algorithm (CS), the result is cuckoo search improved control output
hidden layer algorithm (CS-IM-COH). This new algorithm has a
better ability in optimizing sensitivity of the initial weights and bias
than the original BP algorithm. In this research, the algorithm of
CS-IM-COH is compared with the original BP, the IM-COH, and the
original BP with CS (CS-BP). Furthermore, the selected benchmarks,
four time series samples, are shown in this research for illustration.
The research shows that the CS-IM-COH algorithm give the best
forecasting results compared with the selected samples.
Abstract: Electricity markets throughout the world have
undergone substantial changes. Accurate, reliable, clear and
comprehensible modeling and forecasting of different variables
(loads and prices in the first instance) have achieved increasing
importance. In this paper, we describe the actual state of the
art focusing on reg-SARMA methods, which have proven to be
flexible enough to accommodate the electricity price/load behavior
satisfactory. More specifically, we will discuss: 1) The dichotomy
between point and interval forecasts; 2) The difficult choice between
stochastic (e.g. climatic variation) and non-deterministic predictors
(e.g. calendar variables); 3) The confrontation between modelling
a single aggregate time series or creating separated and potentially
different models of sub-series. The noteworthy point that we would
like to make it emerge is that prices and loads require different
approaches that appear irreconcilable even though must be made
reconcilable for the interests and activities of energy companies.
Abstract: Building systems are highly vulnerable to different
kinds of faults and failures. In fact, various faults, failures and human
behaviors could affect the building performance. This paper tackles
the detection of unreliable sensors in buildings. Different literature
surveys on diagnosis techniques for sensor grids in buildings have
been published but all of them treat only bias and outliers. Occurences
of data gaps have also not been given an adequate span of attention
in the academia. The proposed methodology comprises the automatic thresholding
for data gap detection for a set of heterogeneous sensors in
instrumented buildings. Sensor measurements are considered to be
regular time series. However, in reality, sensor values are not
uniformly sampled. So, the issue to solve is from which delay each
sensor become faulty? The use of time series is required for detection of abnormalities on
the delays. The efficiency of the method is evaluated on measurements
obtained from a real power plant: an office at Grenoble Institute of
technology equipped by 30 sensors.
Abstract: The development of Global Navigation Satellite System (GNSS) technology has led to increasingly widely and successful applications of GNSS surveys for monitoring crustal movements. Instead of the multi-period GNSS solutions, this study utilizes GNSS time series that are required to more precisely determine the vertical deformations in the study area. In recent years, the surface deformations that are parallel and semi-parallel to Bolvadin fault have occurred in Western Anatolia. These surface deformations have continued to occur in Bolvadin settlement area that is located mostly on alluvium ground. Due to these surface deformations, a number of cracks in the buildings located in the residential areas and breaks in underground water and sewage systems have been observed. In order to determine the amount of vertical surface deformations, two continuous GNSS stations have been established in the region. The stations have been operating since 2015 and 2017, respectively. In this study, GNSS observations from the mentioned two GNSS stations were processed with GAMIT/GLOBK (GNSS Analysis Massachusetts Institute of Technology/GLOBal Kalman) program package to create coordinate time series. With the time series analyses, the GNSS stations’ behaviour models (linear, periodical, etc.), the causes of these behaviours, and mathematical models were determined. The study results from the time series analysis of these two 2 GNSS stations show approximately 50-90 mm/yr vertical movement.
Abstract: This study was designed to find the best stochastic model (using of time series analysis) for annual extreme streamflow (peak and maximum streamflow) of Karkheh River at Iran. The Auto-regressive Integrated Moving Average (ARIMA) model used to simulate these series and forecast those in future. For the analysis, annual extreme streamflow data of Jelogir Majin station (above of Karkheh dam reservoir) for the years 1958–2005 were used. A visual inspection of the time plot gives a little increasing trend; therefore, series is not stationary. The stationarity observed in Auto-Correlation Function (ACF) and Partial Auto-Correlation Function (PACF) plots of annual extreme streamflow was removed using first order differencing (d=1) in order to the development of the ARIMA model. Interestingly, the ARIMA(4,1,1) model developed was found to be most suitable for simulating annual extreme streamflow for Karkheh River. The model was found to be appropriate to forecast ten years of annual extreme streamflow and assist decision makers to establish priorities for water demand. The Statistical Analysis System (SAS) and Statistical Package for the Social Sciences (SPSS) codes were used to determinate of the best model for this series.
Abstract: During the last 10 years, in spite of the economic crisis, the number of tourists arriving in Greece has increased, particularly during the tourist season from April to October. In this paper, the number of annual tourist arrivals is studied to explore their preferences with regard to the month of travel, the selected destinations, as well the amount of money spent. The collected data are processed with statistical methods, yielding numerical and graphical results. From the computation of statistical parameters and the forecasting with exponential smoothing, useful conclusions are arrived at that can be used by the Greek tourism authorities, as well as by tourist organizations, for planning purposes for the coming years. The results of this paper and the computed forecast can also be used for decision making by private tourist enterprises that are investing in Greece. With regard to the statistical methods, the method of Simple Exponential Smoothing of time series of data is employed. The search for a best forecast for 2017 and 2018 provides the value of the smoothing coefficient. For all statistical computations and graphics Microsoft Excel is used.
Abstract: Ambient air pollution with fine particulate matter (PM10) is a systematic permanent problem in many countries around the world. The accumulation of a large number of measurements of both the PM10 concentrations and the accompanying atmospheric factors allow for their statistical modeling to detect dependencies and forecast future pollution. This study applies the classification and regression trees (CART) method for building and analyzing PM10 models. In the empirical study, average daily air data for the city of Pleven, Bulgaria for a period of 5 years are used. Predictors in the models are seven meteorological variables, time variables, as well as lagged PM10 variables and some lagged meteorological variables, delayed by 1 or 2 days with respect to the initial time series, respectively. The degree of influence of the predictors in the models is determined. The selected best CART models are used to forecast future PM10 concentrations for two days ahead after the last date in the modeling procedure and show very accurate results.
Abstract: The present study investigates the space-time impact of climate change on the rice crop calendar in tropical Gujranwala, Pakistan. The climate change impact was quantified through the climatic variables, whereas the existing calendar of the rice crop was compared with the phonological stages of the crop, depicted through the time series of the Normalized Difference Vegetation Index (NDVI) derived from Landsat data for the decade 2005-2015. Local maxima were applied on the time series of NDVI to compute the rice phonological stages. Panel models with fixed and cross-section fixed effects were used to establish the relation between the climatic parameters and the time-series of NDVI across villages and across rice growing periods. Results show that the climatic parameters have significant impact on the rice crop calendar. Moreover, the fixed effect model is a significant improvement over cross-sectional fixed effect models (R-squared equal to 0.673 vs. 0.0338). We conclude that high inter-annual variability of climatic variables cause high variability of NDVI, and thus, a shift in the rice crop calendar. Moreover, inter-annual (temporal) variability of the rice crop calendar is high compared to the inter-village (spatial) variability. We suggest the local rice farmers to adapt this change in the rice crop calendar.