Search results for: forecasting model

Authors: Rafid A. Al-Khannak, Fawzi M. Al-Naima

Abstract:

Power load is one of the most important controlling keys which decide power demands and illustrate power usage to shape power market. Hence, power load forecasting is the parameter which facilitates understanding and analyzing all these aspects. In this paper, power load forecasting is solved under MATLAB environment by constructing a neural network for the power load to find an accurate simulated solution with the minimum error. A developed algorithm to achieve load forecasting application with faster technique is the aim for this paper. The algorithm is used to enable MATLAB power application to be implemented by multi machines in the Grid computing system, and to accomplish it within much less time, cost and with high accuracy and quality. Grid Computing, the modern computational distributing technology, has been used to enhance the performance of power applications by utilizing idle and desired Grid contributor(s) by sharing computational power resources.

Keywords: DeskGrid, Grid Server, idle contributor(s), grid computing, load forecasting

Procedia PDF Downloads 449

Authors: Kien-Trinh Thi Bui, Cuong Manh Nguyen

Abstract:

In order to simulate and reproduce the operational characteristics of a dam visually, it is necessary to capture the displacement at different measurement points and analyze the observed movement data promptly to forecast the dam safety. The accuracy of forecasts is further improved by applying machine learning methods to data analysis progress. In this study, the horizontal displacement monitoring data of the Ialy hydroelectric dam was applied to machine learning algorithms: Gaussian processes, multi-layer perceptron neural networks, and the M5-rules algorithm for modelling and forecasting of horizontal displacement of the Ialy hydropower dam (Vietnam), respectively, for analysing. The database which used in this research was built by collecting time series of data from 2006 to 2021 and divided into two parts: training dataset and validating dataset. The final results show all three algorithms have high performance for both training and model validation, but the MLPs is the best model. The usability of them are further investigated by comparison with a benchmark models created by multi-linear regression. The result show the performance which obtained from all the GP model, the MLPs model and the M5-Rules model are much better, therefore these three models should be used to analyze and predict the horizontal displacement of the dam.

Keywords: Gaussian processes, horizontal displacement, hydropower dam, Ialy dam, M5-Rules, multi-layer perception neural networks

Procedia PDF Downloads 176

Authors: Fazıl Gökgöz, Fahrettin Filiz

Abstract:

Load forecasting has become crucial in recent years and become popular in forecasting area. Many different power forecasting models have been tried out for this purpose. Electricity load forecasting is necessary for energy policies, healthy and reliable grid systems. Effective power forecasting of renewable energy load leads the decision makers to minimize the costs of electric utilities and power plants. Forecasting tools are required that can be used to predict how much renewable energy can be utilized. The purpose of this study is to explore the effectiveness of LSTM-based neural networks for estimating renewable energy loads. In this study, we present models for predicting renewable energy loads based on deep neural networks, especially the Long Term Memory (LSTM) algorithms. Deep learning allows multiple layers of models to learn representation of data. LSTM algorithms are able to store information for long periods of time. Deep learning models have recently been used to forecast the renewable energy sources such as predicting wind and solar energy power. Historical load and weather information represent the most important variables for the inputs within the power forecasting models. The dataset contained power consumption measurements are gathered between January 2016 and December 2017 with one-hour resolution. Models use publicly available data from the Turkish Renewable Energy Resources Support Mechanism. Forecasting studies have been carried out with these data via deep neural networks approach including LSTM technique for Turkish electricity markets. 432 different models are created by changing layers cell count and dropout. The adaptive moment estimation (ADAM) algorithm is used for training as a gradient-based optimizer instead of SGD (stochastic gradient). ADAM performed better than SGD in terms of faster convergence and lower error rates. Models performance is compared according to MAE (Mean Absolute Error) and MSE (Mean Squared Error). Best five MAE results out of 432 tested models are 0.66, 0.74, 0.85 and 1.09. The forecasting performance of the proposed LSTM models gives successful results compared to literature searches.

Keywords: deep learning, long short term memory, energy, renewable energy load forecasting

Procedia PDF Downloads 236

Authors: Stephane Laussac, Jacques Piazzola, Gilles Tedeschi

Abstract:

Estimation of the impact of atmospheric aerosols on the climate evolution is an important scientific challenge. One of a major source of particles is constituted by the oceans through the generation of sea-spray aerosols. In coastal areas, marine aerosols can affect air quality through their ability to interact chemically and physically with other aerosol species and gases. The integration of accurate sea-spray emission terms in modeling studies is then required. However, it was found that sea-spray concentrations are not represented with the necessary accuracy in some situations, more particularly at short fetch. In this study, the WRF-Chem model was implemented on a North-Western Mediterranean coastal region. WRF-Chem is the Weather Research and Forecasting (WRF) model online-coupled with chemistry for investigation of regional-scale air quality which simulates the emission, transport, mixing, and chemical transformation of trace gases and aerosols simultaneously with the meteorology. One of the objectives was to test the ability of the WRF-Chem model to represent the fine details of the coastal geography to provide accurate predictions of sea spray evolution for different fetches and the anthropogenic aerosols. To assess the performance of the model, a comparison between the model predictions using a local emission inventory and the physicochemical analysis of aerosol concentrations measured for different wind direction on the island of Porquerolles located 10 km south of the French Riviera is proposed.

Keywords: sea-spray aerosols, coastal areas, sea-spray concentrations, short fetch, WRF-Chem model

Procedia PDF Downloads 171

Authors: Yousef Abu Nahleh, Arun Kumar, Fugen Daver, Reham Al-Hindawi

Abstract:

Despite the availability of natural disaster related time series data for last 110 years, there is no forecasting tool available to humanitarian relief organizations to determine forecasts for emergency logistics planning. This study develops a forecasting tool based on identifying probability of disaster for each country in the world by using decision tree modeling. Further, the determination of aggregate forecasts leads to efficient pre-disaster planning. Based on the research findings, the relief agencies can optimize the various resources allocation in emergency logistics planning.

Keywords: decision tree modeling, forecasting, humanitarian relief, emergency supply chain

Procedia PDF Downloads 456

Authors: Yangrae Cho, Jinseok Kim, Yongtae Park

Abstract:

Mobile app service ecosystem has been abruptly emerged, explosively grown, and dynamically transformed. In contrast with product markets in which product sales directly cause increment in firm’s income, customer’s usage is less visible but more valuable in service market. Especially, the market situation with cutthroat competition in mobile app store makes securing and keeping of users as vital. Although a few service firms try to manage their apps’ usage patterns by fitting on S-curve or applying other forecasting techniques, the time series approaches based on past sequential data are subject to fundamental limitation in the market where customer’s attention is being moved unpredictably and dynamically. We therefore propose a new conceptual approach for detecting usage pattern of mobile app service with Hidden Markov Model (HMM) which is based on the dual stochastic structure and mainly used to clarify unpredictable and dynamic sequential patterns in voice recognition or stock forecasting. Our approach could be practically utilized for app service firms to manage their services’ lifecycles and academically expanded to other markets.

Keywords: mobile app service, usage pattern, Hidden Markov Model, pattern detection

Procedia PDF Downloads 308

Authors: Bandreddy Anand Babu, Srinivasa Rao Mandadi, C. Pradeep Reddy, N. Ramesh Babu

Abstract:

In this paper quickly illustrate the correlation investigation of Auto-Regressive Integrated Moving and Average (ARIMA) and daptive Network Based Fuzzy Inference System (ANFIS) models done by climate estimating. The climate determining is taken from University of Waterloo. The information is taken as Relative Humidity, Ambient Air Temperature, Barometric Pressure and Wind Direction utilized within this paper. The paper is carried out by analyzing the exhibitions are seen by demonstrating of ARIMA and ANIFIS model like with Sum of average of errors. Versatile Network Based Fuzzy Inference System (ANFIS) demonstrating is carried out by Mat lab programming and Auto-Regressive Integrated Moving and Average (ARIMA) displaying is produced by utilizing XLSTAT programming. ANFIS is carried out in Fuzzy Logic Toolbox in Mat Lab programming.

Keywords: ARIMA, ANFIS, fuzzy surmising tool stash, weather forecasting, MATLAB

Procedia PDF Downloads 386

Authors: Fauzia Dianawati, Riska W. Purnomo

Abstract:

Food and agricultural sector become the biggest sector contributing to inflation in Indonesia. Especially in Wonogiri district, red cayenne pepper was the biggest sector contributing to inflation on 2016. A national statistic proved that in recent five years red cayenne pepper has the highest average level of fluctuation among all commodities. Some factors, like supply chain, price disparity, production quantity, crop failure, and oil price become the possible factor causes high volatility level in red cayenne pepper price. Therefore, this research tries to find the key factor causing fluctuation on red cayenne pepper by using ARCH/GARCH method. The method could accommodate the presence of heteroscedasticity in time series data. At the end of the research, it is statistically found that the second level of supply chain becomes the biggest part contributing to inflation with 3,35 of coefficient in fluctuation forecasting model of red cayenne pepper price. This model could become a reference to the government to determine the appropriate policy in maintaining the price stability of red cayenne pepper.

Keywords: ARCH/GARCH, forecasting, red cayenne pepper, volatility, supply chain

Procedia PDF Downloads 164

Authors: Piotr Fiszeder, Marcin Fałdziński, Peter Molnár

Abstract:

The dynamic conditional correlation model of Engle (2002) is one of the most popular multivariate volatility models. However, this model is based solely on closing prices. It has been documented in the literature that the high and low price of the day can be used in an efficient volatility estimation. We, therefore, suggest a model which incorporates high and low prices into the dynamic conditional correlation framework. Empirical evaluation of this model is conducted on three datasets: currencies, stocks, and commodity exchange-traded funds. The utilisation of realized variances and covariances as proxies for true variances and covariances allows us to reach a strong conclusion that our model outperforms not only the standard dynamic conditional correlation model but also a competing range-based dynamic conditional correlation model.

Keywords: volatility, DCC model, high and low prices, range-based models, covariance forecasting

Procedia PDF Downloads 153

Authors: Jeffrey Tzu-Hao Tsai, Yi-Shan Wong

Abstract:

This article proposes a cohort-age-period (CAP) model to characterize multi-population mortality processes using cohort, age, and period variables. Distinct from the factor-based Lee-Carter-type decomposition mortality model, this approach is empirically based and includes the age, period, and cohort variables into the equation system. The model not only provides a fruitful intuition for explaining multivariate mortality change rates but also has a better performance in forecasting future patterns. Using the US and the UK mortality data and performing ten-year out-of-sample tests, our approach shows smaller mean square errors in both countries compared to the models in the literature.

Keywords: longevity risk, stochastic mortality model, multivariate mortality rate, risk management

Procedia PDF Downloads 28

Authors: Kurias George, Ildefonso Cuesta Romeo, Clara Salueña Pérez, Jordi Sole Olle

Abstract:

With the growing prevalence of wind energy there is a need, for modeling techniques to evaluate the impact of wind farms on meteorology and oceanography. This study presents an approach that utilizes the WRF (Weather Research and Forecasting )with that include a Wind Farm Parametrization model to simulate the dynamics around Parc Tramuntana project, a offshore wind farm to be located near the Gulf of Roses off the coast of Barcelona, Catalonia. The model incorporates parameterizations for wind turbines enabling a representation of the wind field and how it interacts with the infrastructure of the wind farm. Current results demonstrate that the model effectively captures variations in temeperature, pressure and in both wind speed and direction over time along with their resulting effects on power output from the wind farm. These findings are crucial for optimizing turbine placement and operation thus improving efficiency and sustainability of the wind farm. In addition to focusing on atmospheric interactions, this study delves into the wake effects within the turbines in the farm. A range of meteorological parameters were also considered to offer a comprehensive understanding of the farm's microclimate. The model was tested under different horizontal resolutions and farm layouts to scrutinize the wind farm's effects more closely. These experimental configurations allow for a nuanced understanding of how turbine wakes interact with each other and with the broader atmospheric and oceanic conditions. This modified approach serves as a potent tool for stakeholders in renewable energy, environmental protection, and marine spatial planning. environmental protection and marine spatial planning. It provides a range of information regarding the environmental and socio economic impacts of offshore wind energy projects.

Keywords: weather research and forecasting, wind turbine wake effects, environmental impact, wind farm parametrization, sustainability analysis

Procedia PDF Downloads 44

Authors: Nato Kutaladze, George Mikuchadze, Giorgi Sokhadze

Abstract:

Precipitation forecasts are one of the most demanding applications in numerical weather prediction (NWP). Georgia, as the whole Caucasian region, is characterized by very complex topography. The country territory is prone to flash floods and mudflows, quantitative precipitation estimation (QPE) and quantitative precipitation forecast (QPF) at any leading time are very important for Georgia. In this study, advanced research weather forecasting model’s skill in QPF is investigated over Georgia’s territory. We have analyzed several convection parameterization and microphysical scheme combinations for different rainy episodes and heavy rainy phenomena. We estimate errors and biases in accumulated 6 h precipitation using different spatial resolution during model performance verification for 12-hour and 24-hour lead time against corresponding rain gouge observations and satellite data. Various statistical parameters have been calculated for the 8-month comparison period, and some skills of model simulation have been evaluated. Our focus is on the formation and organization of convective precipitation systems in a low-mountain region. Several problems in connection with QPF have been identified for mountain regions, which include the overestimation and underestimation of precipitation on the windward and lee side of the mountains, respectively, and a phase error in the diurnal cycle of precipitation leading to the onset of convective precipitation in model forecasts several hours too early.

Keywords: extremal dependence index, false alarm, numerical weather prediction, quantitative precipitation forecasting

Procedia PDF Downloads 123

Authors: Chunhua Liao, Jinfei Wang, Bo Shan, Yang Song, Yongjun He, Taifeng Dong

Abstract:

Winter wheat is one of the main crops in Canada. Forecasting of within-field variability of yield in winter wheat at the early stages is essential for precision farming. However, the crop yield modelling based on high spatial resolution satellite data is generally affected by the lack of continuous satellite observations, resulting in reducing the generalization ability of the models and increasing the difficulty of crop yield forecasting at the early stages. In this study, the correlations between Sentinel-2 data (vegetation indices and reflectance) and yield data collected by combine harvester were investigated and a generalized multivariate linear regression (MLR) model was built and tested with data acquired in different years. It was found that the four-band reflectance (blue, green, red, near-infrared) performed better than their vegetation indices (NDVI, EVI, WDRVI and OSAVI) in wheat yield prediction. The optimum phenological stage for wheat yield prediction with highest accuracy was at the growing stages from the end of the flowering to the beginning of the filling stage. The best MLR model was therefore built to predict wheat yield before harvest using Sentinel-2 data acquired at the end of the flowering stage. Further, to improve the ability of the yield prediction at the early stages, three simple unsupervised domain adaptation (DA) methods were adopted to transform the reflectance data at the early stages to the optimum phenological stage. The winter wheat yield prediction using multiple vegetation indices showed higher accuracy than using single vegetation index. The optimum stage for winter wheat yield forecasting varied with different fields when using vegetation indices, while it was consistent when using multispectral reflectance and the optimum stage for winter wheat yield prediction was at the end of flowering stage. The average testing RMSE of the MLR model at the end of the flowering stage was 604.48 kg/ha. Near the booting stage, the average testing RMSE of yield prediction using the best MLR was reduced to 799.18 kg/ha when applying the mean matching domain adaptation approach to transform the data to the target domain (at the end of the flowering) compared to that using the original data based on the models developed at the booting stage directly (“MLR at the early stage”) (RMSE =1140.64 kg/ha). This study demonstrated that the simple mean matching (MM) performed better than other DA methods and it was found that “DA then MLR at the optimum stage” performed better than “MLR directly at the early stages” for winter wheat yield forecasting at the early stages. The results indicated that the DA had a great potential in near real-time crop yield forecasting at the early stages. This study indicated that the simple domain adaptation methods had a great potential in crop yield prediction at the early stages using remote sensing data.

Keywords: wheat yield prediction, domain adaptation, Sentinel-2, within-field scale

Procedia PDF Downloads 39

Authors: Maria C. Mariani, Md Al Masum Bhuiyan, Osei K. Tweneboah, Hector G. Huizar

Abstract:

This work is devoted to the study of modeling geophysical time series. A stochastic technique with time-varying parameters is used to forecast the volatility of data arising in geophysics. In this study, the volatility is defined as a logarithmic first-order autoregressive process. We observe that the inclusion of log-volatility into the time-varying parameter estimation significantly improves forecasting which is facilitated via maximum likelihood estimation. This allows us to conclude that the estimation algorithm for the corresponding one-step-ahead suggested volatility (with ±2 standard prediction errors) is very feasible since it possesses good convergence properties.

Keywords: Augmented Dickey Fuller Test, geophysical time series, maximum likelihood estimation, stochastic volatility model

Procedia PDF Downloads 276

Authors: Juliana Barcelos Cordeiro, Khashayar Mahani, Farbod Farzan, Mohsen A. Jafari

Abstract:

Energy disaggregation has been focused by many energy companies since energy efficiency can be achieved when the breakdown of energy consumption is known. Companies have been investing in technologies to come up with software and/or hardware solutions that can provide this type of information to the consumer. On the other hand, not all people can afford to have these technologies. Therefore, in this paper, we present a methodology for breaking down the aggregate consumption and identifying the highdemanding end-uses profiles. These energy profiles will be used to build the forecast model for optimal control purpose. A facility with high cooling load is used as an illustrative case study to demonstrate the results of proposed methodology. We apply a high level energy disaggregation through a pattern recognition approach in order to extract the consumption profile of its rooftop packaged units (RTUs) and present a forecast model for the energy consumption.  

Keywords: energy consumption forecasting, energy efficiency, load disaggregation, pattern recognition approach

Procedia PDF Downloads 248

Authors: Benny Benyamin Nasution, Rahmat Widia Sembiring, Abdul Rahman Dalimunthe, Nursiah Mustari, Nisfan Bahri, Berta br Ginting, Riadil Akhir Lubis, Rita Tavip Megawati, Indri Dithisari

Abstract:

The earth coordinate system is an important part of an attempt for earthquake forecasting, such as the one using Single Layer Hierarchical Graph Neuron (SLHGN). However, there are a number of problems that need to be worked out before the coordinate system can be utilized for the forecaster. One example of those is that SLHGN requires that the focused area of an earthquake must be constructed in a grid-like form. In fact, within the current earth coordinate system, the same longitude-difference would produce different distances. This can be observed at the distance on the Equator compared to distance at both poles. To deal with such a problem, a coordinate system has been developed, so that it can be used to support the ongoing earthquake forecasting using SLHGN. Two important issues have been developed in this system: 1) each location is not represented through two-value (longitude and latitude), but only a single value, 2) the conversion of the earth coordinate system to the x-y cartesian system requires no angular formulas, which is therefore fast. The accuracy and the performance have not been measured yet, since earthquake data is difficult to obtain. However, the characteristics of the SLHGN results show a very promising answer.

Keywords: hierarchical graph neuron, multidimensional hierarchical graph neuron, single layer hierarchical graph neuron, natural disaster forecasting, earthquake forecasting, earth coordinate system

Procedia PDF Downloads 196

Authors: Etienne Gaborit, Dorothy Durnford, Daniel Deacu, Marco Carrera, Nathalie Gauthier, Camille Garnaud, Vincent Fortin

Abstract:

A new experimental streamflow forecasting system was recently implemented at the Environment and Climate Change Canada’s (ECCC) Canadian Centre for Meteorological and Environmental Prediction (CCMEP). It relies on CaLDAS (Canadian Land Data Assimilation System) for the assimilation of surface variables, and on a surface prediction system that feeds a routing component. The surface energy and water budgets are simulated with the SVS (Soil, Vegetation, and Snow) Land-Surface Scheme (LSS) at 2.5-km grid spacing over Canada. The routing component is based on the Watroute routing scheme at 1-km grid spacing for the Great Lakes and Nelson River watersheds. The system is run in two distinct phases: an analysis part and a forecast part. During the analysis part, CaLDAS outputs are used to force the routing system, which performs streamflow assimilation. In forecast mode, the surface component is forced with the Canadian GEM atmospheric forecasts and is initialized with a CaLDAS analysis. Streamflow performances of this new system are presented over 2019. Performances are compared to the current ECCC’s operational streamflow forecasting system, which is different from the new experimental system in many aspects. These new streamflow forecasts are also compared to persistence. Overall, the new streamflow forecasting system presents promising results, highlighting the need for an elaborated assimilation phase before performing the forecasts. However, the system is still experimental and is continuously being improved. Some major recent improvements are presented here and include, for example, the assimilation of snow cover data from remote sensing, a backward propagation of assimilated flow observations, a new numerical scheme for the routing component, and a new reservoir model.

Keywords: assimilation system, distributed physical model, offline hydro-meteorological chain, short-term streamflow forecasts

Procedia PDF Downloads 112

Authors: Qais H. Alsafasfeh

Abstract:

The aim of this paper is to provide a forecast of the peak demand for the next 15 years for electrical distribution companies. The proposed methodology provides both the peak demand and its location for the next 15 years. This paper describes the Spatial Load Forecasting model used, the information provided by electrical distribution company in Jordan, the workflow followed, the parameters used and the assumptions made to run the model. The aim of this paper is to provide a forecast of the peak demand for the next 15 years for electrical distribution companies. The proposed methodology provides both the peak demand and its location for the next 15 years. This paper describes the Spatial Load Forecasting model used, the information provided by electrical distribution company in Jordan, the workflow followed, the parameters used and the assumptions made to run the model.

Keywords: load forecast, peak demand, spatial load, electrical distribution

Procedia PDF Downloads 467

Authors: Flavio Cordeiro Fontanella, Asla Medeiros e Sá, Moacyr Alvim Horta Barbosa da Silva

Abstract:

In the present work, we analyse the performance of football results forecasting models. In order to do so, we have performed the data collection from eight different forecasting models during the 2021 Brazilian football season. First, we guide the analysis through visual representations of the data, designed to highlight the most prominent features and enhance the interpretation of differences and similarities between the models. We propose using a 2-simplex triangle to investigate visual patterns from the results forecasting models. Next, we compute the expected points for every team playing in the championship and compare them to the final league standings, revealing interesting contrasts between actual to expected performances. Then, we evaluate forecasts’ accuracy using the Ranked Probability Score (RPS); models comparison accounts for tiny scale differences that may become consistent in time. Finally, we observe that the Wisdom of Crowds principle can be appropriately applied in the context, driving into a discussion of results forecasts usage in practice. This paper’s primary goal is to encourage football forecasts’ performance discussion. We hope to accomplish it by presenting appropriate criteria and easy-to-understand visual representations that can point out the relevant factors of the subject.

Keywords: accuracy evaluation, Brazilian championship, football results forecasts, forecasting models, visual analysis

Procedia PDF Downloads 69

Authors: Niyati Sharma, Om Anand, Sanjeev Kumar Prasad

Abstract:

Assuming future sale value for any of the organizations is one of the major essential characteristics of tactical development. Walmart Sales Forecasting is the finest illustration to work with as a beginner; subsequently, it has the major retail data set. Walmart uses this sales estimate problem for hiring purposes also. We would like to analyzing how the internal and external effects of one of the largest companies in the US can walk out their Weekly Sales in the future. Demand forecasting is the planned prerequisite of products or services in the imminent on the basis of present and previous data and different stages of the market. Since all associations is facing the anonymous future and we do not distinguish in the future good demand. Hence, through exploring former statistics and recent market statistics, we envisage the forthcoming claim and building of individual goods, which are extra challenging in the near future. As a result of this, we are producing the required products in pursuance of the petition of the souk in advance. We will be using several machine learning models to test the exactness and then lastly, train the whole data by Using linear regression and fitting the training data into it. Accuracy is 8.88%. The extra trees regression model gives the best accuracy of 97.15%.

Keywords: random forest algorithm, linear regression algorithm, extra trees classifier, mean absolute error

Procedia PDF Downloads 121

Authors: Yuzhi Cai

Abstract:

In this paper, we introduce a newly developed quantile function model that can be used for estimating conditional distributions of financial returns and for obtaining multi-step ahead out-of-sample predictive distributions of financial returns. Since we forecast the whole conditional distributions, any predictive quantity of interest about the future financial returns can be obtained simply as a by-product of the method. We also show an application of the model to the daily closing prices of Dow Jones Industrial Average (DJIA) series over the period from 2 January 2004 - 8 October 2010. We obtained the predictive distributions up to 15 days ahead for the DJIA returns, which were further compared with the actually observed returns and those predicted from an AR-GARCH model. The results show that the new model can capture the main features of financial returns and provide a better fitted model together with improved mean forecasts compared with conventional methods. We hope this talk will help audience to see that this new model has the potential to be very useful in practice.

Keywords: DJIA, financial returns, predictive distribution, quantile function model

Procedia PDF Downloads 346

Authors: Jonathan Iworiso

Abstract:

Forecasting the equity premium out-of-sample is a major concern to researchers in finance and emerging markets. The quest for a superior model that can forecast the equity premium with significant economic gains has resulted in several controversies on the choice of variables and suitable techniques among scholars. This research focuses mainly on the application of Regression Training (RT) techniques to forecast monthly equity premium out-of-sample recursively with an expanding window method. A broad category of sophisticated regression models involving model complexity was employed. The RT models include Ridge, Forward-Backward (FOBA) Ridge, Least Absolute Shrinkage and Selection Operator (LASSO), Relaxed LASSO, Elastic Net, and Least Angle Regression were trained and used to forecast the equity premium out-of-sample. In this study, the empirical investigation of the RT models demonstrates significant evidence of equity premium predictability both statistically and economically relative to the benchmark historical average, delivering significant utility gains. They seek to provide meaningful economic information on mean-variance portfolio investment for investors who are timing the market to earn future gains at minimal risk. Thus, the forecasting models appeared to guarantee an investor in a market setting who optimally reallocates a monthly portfolio between equities and risk-free treasury bills using equity premium forecasts at minimal risk.

Keywords: regression training, out-of-sample forecasts, expanding window, statistical predictability, economic significance, utility gains

Procedia PDF Downloads 75

Authors: Nicola Rubino

Abstract:

This paper presents an analysis of a group of commodity exporting countries' nominal exchange rate movements in relationship to the US dollar. Using a series of Unrestricted Self-exciting Threshold Autoregressive models (SETAR), we model and evaluate sixteen national CPI price differentials relative to the US dollar CPI. Out-of-sample forecast accuracy is evaluated through calculation of mean absolute error measures on the basis of two-hundred and fifty-three months rolling window forecasts and extended to three additional models, namely a logistic smooth transition regression (LSTAR), an additive non linear autoregressive model (AAR) and a simple linear Neural Network model (NNET). Our preliminary results confirm presence of some form of TAR non linearity in the majority of the countries analyzed, with a relatively higher goodness of fit, with respect to the linear AR(1) benchmark, in five countries out of sixteen considered. Although no model appears to statistically prevail over the other, our final out-of-sample forecast exercise shows that SETAR models tend to have quite poor relative forecasting performance, especially when compared to alternative non-linear specifications. Finally, by analyzing the implied half-lives of the > coefficients, our results confirms the presence, in the spirit of arbitrage band adjustment, of band convergence with an inner unit root behaviour in five of the sixteen countries analyzed.

Keywords: transition regression model, real exchange rate, nonlinearities, price differentials, PPP, commodity points

Procedia PDF Downloads 255

Authors: Giovanni Masala, Stefania Marica

Abstract:

Forecasting electricity load plays a crucial role regards decision making and planning for economical purposes. Besides, in the light of the recent privatization and deregulation of the power industry, the forecasting of future electricity load turned out to be a very challenging problem. Empirical data about electricity load highlights a clear seasonal behavior (higher load during the winter season), which is partly due to climatic effects. We also emphasize the presence of load periodicity at a weekly basis (electricity load is usually lower on weekends or holidays) and at daily basis (electricity load is clearly influenced by the hour). Finally, a long-term trend may depend on the general economic situation (for example, industrial production affects electricity load). All these features must be captured by the model. The purpose of this paper is then to build an hourly electricity load model. The deterministic component of the model requires non-linear regression and Fourier series while we will investigate the stochastic component through econometrical tools. The calibration of the parameters’ model will be performed by using data coming from the Italian market in a 6 year period (2007- 2012). Then, we will perform a Monte Carlo simulation in order to compare the simulated data respect to the real data (both in-sample and out-of-sample inspection). The reliability of the model will be deduced thanks to standard tests which highlight a good fitting of the simulated values.

Keywords: ARMA-GARCH process, electricity load, fitting tests, Fourier series, Monte Carlo simulation, non-linear regression

Procedia PDF Downloads 377

Authors: F. Lazzeri, I. Reiter

Abstract:

Energy production optimization has been traditionally very important for utilities in order to improve resource consumption. However, load forecasting is a challenging task, as there are a large number of relevant variables that must be considered, and several strategies have been used to deal with this complex problem. This is especially true also in microgrids where many elements have to adjust their performance depending on the future generation and consumption conditions. The goal of this paper is to present a solution for short-term load forecasting in microgrids, based on three machine learning experiments developed in R and web services built and deployed with different components of Cortana Intelligence Suite: Azure Machine Learning, a fully managed cloud service that enables to easily build, deploy, and share predictive analytics solutions; SQL database, a Microsoft database service for app developers; and PowerBI, a suite of business analytics tools to analyze data and share insights. Our results show that Boosted Decision Tree and Fast Forest Quantile regression methods can be very useful to predict hourly short-term consumption in microgrids; moreover, we found that for these types of forecasting models, weather data (temperature, wind, humidity and dew point) can play a crucial role in improving the accuracy of the forecasting solution. Data cleaning and feature engineering methods performed in R and different types of machine learning algorithms (Boosted Decision Tree, Fast Forest Quantile and ARIMA) will be presented, and results and performance metrics discussed.

Keywords: time-series, features engineering methods for forecasting, energy demand forecasting, Azure Machine Learning

Procedia PDF Downloads 276

Authors: Sibel Celik, Hüseyin Ergin

Abstract:

The paper aims to find the best volatility forecasting model for stock markets in Turkey. For this purpose, we compare performance of different volatility models-both traditional GARCH model and high frequency based volatility models- and conclude that both in pre-crisis and crisis period, the performance of high frequency based volatility models are better than traditional GARCH model. The findings of paper are important for policy makers, financial institutions and investors.

Keywords: volatility, GARCH model, realized volatility, high frequency data

Procedia PDF Downloads 469

Authors: R. Premasiri, W. A. H. A. Abeygunasekara, S. M. Hewavidana, T. Jananthan, R. M. S. Madawala, K. Vaheeshan

Abstract:

Forecasting the potential for disastrous events such as landslides has become one of the major necessities in the current world. Most of all, the landslides occurred in Sri Lanka are found to be triggered mostly by intense rainfall events. The study area is the landslide near Gerandiella waterfall which is located by the 41st kilometer post on Nuwara Eliya-Gampala main road in Kotmale Division in Sri Lanka. The landslide endangers the entire Kotmale town beneath the slope. Geographic Information System (GIS) platform is very much useful when it comes to the need of emulating the real-world processes. The models are used in a wide array of applications ranging from simple evaluations to the levels of forecast future events. This project investigates the possibility of developing a dynamic model to map the spatial distribution of the slope stability. The model incorporates several theoretical models including the infinite slope model, Green Ampt infiltration model and Perched ground water flow model. A series of rainfall values can be fed to the model as the main input to simulate the dynamics of slope stability. Hydrological model developed using GIS is used to quantify the perched water table height, which is one of the most critical parameters affecting the slope stability. Infinite slope stability model is used to quantify the degree of slope stability in terms of factor of safety. DEM was built with the use of digitized contour data. Stratigraphy was modeled in Surfer using borehole data and resistivity images. Data available from rainfall gauges and piezometers were used in calibrating the model. During the calibration, the parameters were adjusted until a good fit between the simulated ground water levels and the piezometer readings was obtained. This model equipped with the predicted rainfall values can be used to forecast of the slope dynamics of the area of interest. Therefore it can be investigated the slope stability of rainfall induced landslides by adjusting temporal dimensions.

Keywords: factor of safety, geographic information system, hydrological model, slope stability

Procedia PDF Downloads 393

Authors: Yunwei Zhang, Qiaochu Tang, Yao Zhang, Jiabin Wang, Ulrich Stimming, Alpha Lee

Abstract:

Forecasting the state of health and remaining useful life of Li-ion batteries is an unsolved challenge that limits technologies such as consumer electronics and electric vehicles. Here we build an accurate battery forecasting system by combining electrochemical impedance spectroscopy (EIS) -- a real-time, non-invasive and information-rich measurement that is hitherto underused in battery diagnosis -- with Gaussian process machine learning. We collect over 20,000 EIS spectra of commercial Li-ion batteries at different states of health, states of charge and temperatures -- the largest dataset to our knowledge of its kind. Our Gaussian process model takes the entire spectrum as input, without further feature engineering, and automatically determines which spectral features predict degradation. Our model accurately predicts the remaining useful life, even without complete knowledge of past operating conditions of the battery. Our results demonstrate the value of EIS signals in battery management systems.

Keywords: battery degradation, machine learning method, electrochemical impedance spectroscopy, battery diagnosis

Procedia PDF Downloads 114

Authors: Eyad Almaita

Abstract:

In this paper, a novel adaptive Radial Basis Function Neural Networks (RBFNN) algorithm is used to forecast the hour by hour electrical load demand in Jordan. A small and effective RBFNN model is used to forecast the hourly total load demand based on a small number of features. These features are; the load in the previous day, the load in the same day in the previous week, the temperature in the same hour, the hour number, the day number, and the day type. The proposed adaptive RBFNN model can enhance the reliability of the conventional RBFNN after embedding the network in the system. This is achieved by introducing an adaptive algorithm that allows the change of the weights of the RBFNN after the training process is completed, which will eliminates the need to retrain the RBFNN model again. The data used in this paper is real data measured by National Electrical Power co. (Jordan). The data for the period Jan./2012-April/2013 is used train the RBFNN models and the data for the period May/2013- Sep. /2013 is used to validate the models effectiveness.

Keywords: load forecasting, adaptive neural network, radial basis function, short-term, electricity consumption

Procedia PDF Downloads 318

Authors: Ulrike Dowie, Ralph Grothmann

Abstract:

Rational planning decisions are based upon forecasts. Precise forecasting has, therefore, a central role in business. The prediction of customer demand is a prime example. This paper introduces recurrent neural networks to model customer demand and combines the forecast with uncertainty measures to derive decision support of the demand planning department. It identifies and describes the keys to the successful implementation of an AI-based solution: bringing together data with business knowledge, AI methods, and user experience, and applying agile software development practices.

Keywords: agile software development, AI project success factors, deep learning, demand forecasting, forecast uncertainty, neural networks, supply chain management

Procedia PDF Downloads 148