Classical Time Series Research Articles

Comparative Analysis of Methods for Hourly Electricity Demand Forecasting in the Absence of Data – A Case Study

Abstract Scope and purpose of work This paper examines the impact of the number of gaps in data, the analytical form, and the model type selection criterion on the accuracy of interpolation and extrapolation forecasts for hourly data. Materials and methods Forecasts were developed on the basis of predictors that are based on: classical time series forecasting models and regression time series forecasting models, hybrid time series forecasting models and hybrid regression forecasting models for uncleared series, and exponential smoothing models for cleared series of two or three types of seasonal fluctuations, with minimum estimates of errors in interpolation or extrapolation forecasts. Results Adaptive and hybrid regression models have proved to have the most favorable predictive properties. Most hybrid time series models for systematic and non-systematic gaps and for both analytical forms are single models that generally describe fluctuations within a 24-hour cycle. Conclusions The lowest estimators of prediction errors involving interpolation were obtained for exponential smoothing models, followed by hybrid regression models. A reverse sequence was obtained for extrapolative forecasting.

Forecasting Long-Term Electricity Consumption in Saudi Arabia Based on Statistical and Machine Learning Algorithms to Enhance Electric Power Supply Management

This study aims to develop statistical and machine learning methodologies for forecasting yearly electricity consumption in Saudi Arabia. The novelty of this study include (i) determining significant features that have a considerable influence on electricity consumption, (ii) utilizing a Bayesian optimization algorithm (BOA) to enhance the model’s hyperparameters, (iii) hybridizing the BOA with the machine learning algorithms, viz., support vector regression (SVR) and nonlinear autoregressive networks with exogenous inputs (NARX), for modeling individually the long-term electricity consumption, (iv) comparing their performances with the widely used classical time-series algorithm autoregressive integrated moving average with exogenous inputs (ARIMAX) with regard to the accuracy, computational efficiency, and generalizability, and (v) forecasting future yearly electricity consumption and validation. The population, gross domestic product (GDP), imports, and refined oil products were observed to be significant with the total yearly electricity consumption in Saudi Arabia. The coefficient of determination R2 values for all the developed models are >0.98, indicating an excellent fit of the models with historical data. However, among all three proposed models, the BOA–NARX has the best performance, improving the forecasting accuracy (root mean square error (RMSE)) by 71% and 80% compared to the ARIMAX and BOA–SVR models, respectively. The overall results of this study confirm the higher accuracy and reliability of the proposed methods in total electricity consumption forecasting that can be used by power system operators to more accurately forecast electricity consumption to ensure the sustainability of electric energy. This study can also provide significant guidance and helpful insights for researchers to enhance their understanding of crucial research, emerging trends, and new developments in future energy studies.

Comparative evaluation of classical and SARIMA-BL time series hybrid models in predicting monthly qualitative parameters of Maroon river

In the present study, seasonal autoregressive integrated moving average (SARIMA) time series models, nonlinear BL model, multi-layer perceptron artificial neural network and SARIMA-bilinear hybrid models were employed to predict the quality parameters of total dissolved solids (TDS), sodium adsorption ratio (SAR), electrical conductivity (EC) in the Maroon basin in Khuzestan Province. For fitting the mentioned models, the monthly data were used for the calibration (1970–2000), confirmation (2001–2011) and prediction of the model (2012–2018). The appropriate models of SARIMA, the bilinear BL and SARIMA-BL hybrid models of the rationalized parameters of the above-mentioned quality parameters were selected based on the adequacy tests, such as the Akaike criterion and the independence test of the model residuals (Ljung–Box). To determine the effective input parameters of the network, the partial mutual information (PMI) algorithm was used to model the three monthly EC, TDS, and SAR parameters. Also, for input and output layers linear transfer function and for hidden layer various active functions with the back-propagation learning algorithm were used for modeling and predicting this water quality parameters. Comparison of the models showed the tangible superiority of SARIMA-bilinear hybrid models than the artificial neural network with effective input parameters based on PMI algorithm, SARIMA and bl models in performed prediction of the all three monthly qualitative parameters in all three stages, training-validation-test in Maroon basin.

A portmanteau-type test for detecting serial correlation in locally stationary functional time series

The portmanteau test provides the vanilla method for detecting serial correlations in classical univariate time series analysis. The method is extended to the case of observations from a locally stationary functional time series. Asymptotic critical values are obtained by a suitable block multiplier bootstrap procedure. The test is shown to asymptotically hold its level and to be consistent against general alternatives.

THE APPLICATION OF TIME-SERIES FORECASTING MODELS IN GROCERY RETAIL INDUSTRY

The grocery retail landscape in Ukraine has witnessed profound transformations, driven by disruptions like the COVID-19 pandemic and full-scale invasion, leading to unstable consumer behavior and market dynamics. In response, forecasting models must evolve to consider stochastic exogenous factors, such as blackout periods and air alarms. This study explores advanced time series forecasting models and proposes a comprehensive framework for optimal model selection. The study introduces the Neural Prophet, a model that combines interpretability and predictive power by incorporating components like non-periodic trends, periodic seasonality, holiday effects, and regressors. The research methodology involves a comparative analysis of classical time series forecasting methods, machine learning regression approaches, and neural networks. Noteworthy models include LightGBM, RNN, TCN, and Neural Hierarchical Interpolation for Time Series (N-HiTS). Optuna hyperparameter optimization and k-fold cross-validation enhance model accuracy. The study applies the proposed framework to forecast order quantities in the e-commerce segment of the Ukrainian grocery retail company. The system accommodates diverse factors like weather, holidays, and promotions, providing robust decision support. Anomalies are detected using the IQR method, and missing values are filled using Exponentially Weighted Moving Average. Results show the Neural Prophet consistently outperforming other models in 65% of cases, emphasizing its superiority. However, a complete transition to neural models results in reduced accuracy, highlighting the need for a nuanced approach based on data characteristics. The study presents a sophisticated framework for forecasting accuracy, supporting effective operational decision-making. Future research should explore ensemble methods while maintaining computational efficiency, aligning with the ongoing pursuit of optimized forecasting accuracy for informed decision-making in grocery retail.

Comparing classical time-series models and machine learning for demand forecast on the beverage industry in COVID-19 pandemic.

Comparing classical time-series models and machine learning for demand forecast on the beverage industry in COVID-19 pandemic.

A novel evolutionary analysis model for social collaborative design ecosystem based on information entropy

Social collaborative design rooted in online platforms has emerged as a new paradigm for product innovation, with open and dynamic design elements bringing about external design resources and internal fluctuations. To guarantee sustainable operation of the design system, assessing the impact of the uncertainty of the external environment and randomness of the internal organization on the efficiency and stability of the system’s operation is essential. To that end, this paper proposes a quantitative evolutionary analysis model based on information entropy theory to assess the benefits and drawbacks of system evolution by measuring the degree of development and harmony of the system, as well as to anticipate future evolutionary trajectories using a classical time-series prediction model. The accuracy and validity of the model were verified using 794,509 design data points accumulated over 10 years on the Zhubajie platform, the largest social collaborative design platform in China. In summary, our study provides a method for stakeholders to evaluate the evolutionary trend of the design ecosystem. Platform managers can utilize this model for controlling the operating state of the system to ensure safe and effective operation.

Covid-19’un Dünya Finans Piyasaları Üzerindeki Etkisini Belirlemeye Yönelik Bir Analiz

Bu çalışmada, tüm dünyayı etkisine alarak küresel bir salgın haline gelen koronavirüsün, Dünya’nın farklı bölgelerinde bulunan ABD, Almanya, Çin, Japonya, Mısır ve Türkiye’deki borsa endekslerinin volatilitesinde ve likiditelerinde herhangi bir değişiklik oluşturup oluşturmadığı tespit edilmeye çalışılmıştır. Bu çerçevede ABD’de DOW30 Endeksi, Almanya’da DAX Endeksi, Çin’de SSE Bileşik Şangay Endeksi, Japonya’da NİKKEİ 225 Endeksi, Mısır’da EGX30 Endeksi ve Türkiye’de BIST100 Endeksleri üzerinde koronavirüs salgınının etkileri araştırılmıştır. Çalışmanın analiz sonuçları ve yorumlar ilgili ülkelerle sınırlı olup çalışmanın kısıtını oluşturmuştur. Söz konusu ülkeler, gelişmiş ve gelişmekte olan ülkeler arasından ve coğrafi konum açısından içinde bulunduğu bölgeyi temsiliyet gücü yüksek finansal piyasalar olduğu kabul edilerek analize seçilmiştir. Araştırma kapsamında incelenen her ülkede Dünya Sağlık Örgütü (DSÖ)’nün açıkladığı ilk vakanın görüldüğü tarih baz alınmıştır. Her ülke için ilk vakanın görüldüğü tarihten 18 Kasım 2020 tarihine kadarki dönem için veri seti hazırlanmıştır. Pandemi öncesi ve sonrası farklılaşmayı belirleyebilmek için aynı sayıdaki veri kadar pandemi öncesi dönem veri seti de oluşturulmuştur. Böylece pandemi öncesi ve sonrası dönem için bir farklılaşmanın olup olmadığı tespit edilmeye çalışılmıştır. Endekslere ilişkin getiri ve likidite serileri koşullu varyans modellerinden GARCH(1,1) ile tahmin edilmiş ve COVID-19 sonrasında ilgili borsa endekslerinin volatilite ve likiditesinde değişmelerin olduğu gözlemlenmiştir. Araştırmaya konu ülke borsa endekslerin hepsinin getiri serilerinin klasik finansal zaman serilerinde gözlemlenen kalın kuyruk ve çarpıklık özellikleri tespit edilmiştir. Ayrıca volatilite kümelenmeleri gözlemlenmiştir.

Hybrid Models in Forecasting Short Term Road Traffic

Objectives: To analyze the road traffic in Hyderabad at a selected junction and fit Hybrid model for forecasting the traffic conditions for the next near future also to compare the proposed model with the existing popular models and identifying the best model for forecasting the Hyderabad road traffic. Methods: This study considers the data on traffic flow at 6.no. junction in Amberpet, Hyderabad, Telangana state, India. The traffic data has been considered for peak hours in the morning for 8 A.M to 12 Noon, for 6 days. Classical Time series models and recent revolution in time series modelling - Artificial Neural Networks have been applied to develop a forecasting model. Then we combined these models to obtain the Hybrid model for forecasting the road traffic at the selected point. SPSS and Zytun softwares have been used for the analysis. The best fit statistics such as RMSE, MAPE and MAE have been used to identify the best model. Findings: Our study indicates the efficacy of the new combinatorial model in acquiring more accurate forecasting as compared to independent models. These results can be considered to monitor traffic signals and explore methods to avoid congestion at that junction. Novelty: Though some research has been done in Hyderabad, no work has been done to develop a model for forecasting the Hyderabad road traffic. Our model can give the best forecasts for the Hyderabad road traffic. Keywords: Multilayer Perceptron; Hybrid model; Exponential smoothing; Artificial Neural Network; Intelligent Transport System

Küresel iklim değişikliğiyle mücadelede deniz yüzeyi sıcaklığının ileri beslemeli yapay sinir ağları ile tahminlenmesi: Rize örneği, Türkiye

The increase of the world population, especially in the global competition, together with the increasing use of fossil fuel resources to meet energy needs, leads to more greenhouse gases (more than one CO2, methane etc.) emissions and the global climate crisis. In this process, changes in meteorological events such as temperature, precipitation, and wind, attract attention moreover but when considered as a whole, we know that these negative changes in the ecosystem negatively affect many living groups. Sea Surface Temperature (SST) as measured meteorologically is the most important environmental parameter where these changes are monitored and observed. It draws attention to the fact that changes in SST are not limited to living organisms as habitats, but also catalyze many chain reactions, especially socio-economic impacts. Therefore, much of the work is devoted to forecasting studies to adapt to changing habitats and take the necessary precautions against potential risks. Feed-forward artificial neural networks have been commonly used to address these emerging needs. Artificial neural networks, which are a simple imitation of the human neurological system, have been used as an artificial intelligence method in forecasting problems due to their superior performance and not having the limitations of classical time series. In this study, the forecasting of the time series of monthly mean SST temperature obtained from Rize station between the years 2010 and 2020 is performed by using feed-forward artificial neural networks, and the forecasting performance of the corresponding time series is compared with many forecasting methods with different characteristics. The comparison of the methods used the mean square error and mean absolute percentage error criteria, which are commonly used in the forecasting literature. The analysis results showed that the analysis results obtained with the feed-forward artificial neural networks have the best prediction performance. As a result, it can be stated that the sea surface temperature can be forecasted with a very high accuracy using the feed-forward artificial neural networks.

Multivariate Demand Forecasting for Rental Bike Systems Based on an Unobserved Component Model

Many German cities, municipalities and transport associations are expanding their bike-sharing systems (BSS) to offer citizens a cost-effective and climate-friendly means of transport and an alternative to private motorized transport (PMT). However, operators face the challenge of generating high-quality predictive analyses and time series forecasts. In particular, the prediction of demand is a key component to foster data-driven decisions. To address this problem, an Unobserved Component Model (UCM) has been developed to predict the monthly rentals of a BSS, whereby the station-based BSS VRNnextbike, including over 2000 bikes, 297 stations and 21 municipalities, is employed as an example. The model decomposes the time series into trend, seasonal, cyclical, auto-regressive and irregular components for statistical modeling. Additionally, the model includes exogenous factors such as weather, user behavior (e.g., traveled distance), school holidays and COVID-19 relevant covariates as independent effects to calculate scenario based forecasts. It can be shown that the UCM calculates reasonably accurate forecasts and outperforms classical time series models such as ARIMA(X) or SARIMA(X). Improvements were observed in model quality in terms of AIC/BIC (2.5% to 22%) and a reduction in error metrics from 15% to 45% depending on the considered model.

Graphical Assistant Grouped Network Autoregression Model: A Bayesian Nonparametric Recourse

Vector autoregression model is ubiquitous in classical time series data analysis. With the rapid advance of social network sites, time series data over latent graph is becoming increasingly popular. In this article, we develop a novel Bayesian grouped network autoregression model, which can simultaneously estimate group information (number of groups and group configurations) and group-wise parameters. Specifically, a graphically assisted Chinese restaurant process is incorporated under the framework of the network autoregression model to improve the statistical inference performance. An efficient Markov chain Monte Carlo sampling algorithm is used to sample from the posterior distribution. Extensive studies are conducted to evaluate the finite sample performance of our proposed methodology. Additionally, we analyze two real datasets as illustrations of the effectiveness of our approach.

LOCAL INVENTORY DEMAND FORECASTING OF E-COMMERCE WITH MAPREDUCE FRAMEWORK

With the rapid expansion of e-commerce businesses shortest delivery time is a challenging task. As well as price and quality, delivery time is becoming an important key-factor in the race of business growth. To deliver the product to the buyer in the shortest time, the product should be at the local inventory at the time of purchasing. Which is a difficult task for e-commerce companies to store the right products at right time in local inventories. Extracting useful patterns and forecasting demand from purchase history is a tough job because of the huge volume and high variety nature of e-commerce data. The MapReduce algorithm of Big Data ecosystem can be used to pre-process the enormous amount of data and to summarize it into a suitable format which later can easily be used to build any forecasting model. A comparison of different classical time series forecasting methods with the pre-processing of MapReduce algorithm is focused on this work.

Flood River Water Level Forecasting using Ensemble Machine Learning for Early Warning Systems

Flood forecasting is crucial for early warning system and disaster risk reduction. Yet the flood river water levels are difficult and challenging task that it cannot be easily captured with classical time-series approaches. This study proposed a novel intelligence system utilised various machine learning techniques as individual models, including radial basis function neural network (RBF-NN), adaptive neuro-fuzzy inference system (ANFIS), support vector machine (SVM), and long short-term memory network (LSTM) to establish intelligent committee machine learning flood forecasting (ICML-FF) framework. The combination of these individual models achieved through simple averaging method, and further optimised using weighted averaging by K-nearest neighbour (K-NN) and genetic algorithm (GA). The effectiveness of the proposed model was evaluated using real case study for Malaysia’s Kelantan River. The results show that ANFIS outperforms as individual model, while ICML-FF-based model produced better accuracy and lowest error than any one of the individuals. In general, it is found that the proposed ICML-FF is capable of robust forecasting model for flood early warning systems.

Wavelet-based fuzzy clustering of interval time series

We investigate the fuzzy clustering of interval time series using wavelet variances and covariances; in particular, we use a fuzzy c-medoids clustering algorithm. Traditional hierarchical and non-hierarchical clustering methods lead to the identification of mutually exclusive clusters whereas fuzzy clustering methods enable the identification of overlapping clusters, implying that one or more series could belong to more than one cluster simultaneously. An interval time series (ITS) which arises when interval-valued observations are recorded over time is able to capture the variability of values within each interval at each time point. This is in contrast to single-point information available in a classical time series. Our main contribution is that by combining wavelet analysis, interval data analysis and fuzzy clustering, we are able to capture information which would otherwise have not been contemplated by the use of traditional crisp clustering methods on classical time series for which just a single value is recorded at each time point. Through simulation studies, we show that under some circumstances fuzzy c-medoids clustering performs better when applied to ITS than when it is applied to the corresponding traditional time series. Applications to exchange rates ITS and sea-level ITS show that the fuzzy clustering method reveals different and more meaningful results than when applied to associated single-point time series.

Comparing three types of data-driven models for monthly evapotranspiration prediction under heterogeneous climatic conditions

Evapotranspiration is one of the most important hydro-climatological components which directly affects agricultural productions. Therefore, its forecasting is critical for water managers and irrigation planners. In this study, adaptive neuro-fuzzy inference system (ANFIS) model has been hybridized by differential evolution (DE) optimization algorithm as a novel approach to forecast monthly reference evapotranspiration (ET0). Furthermore, this model has been compared with the classic stochastic time series model. For this, the ET0 rates were calculated on a monthly scale during 1995–2018, based on FAO-56 Penman–Monteith equation and meteorological data including minimum air temperature, maximum air temperature, mean air temperature, minimum relative humidity, maximum relative humidity & sunshine duration. The investigation was performed on 6 stations in different climates of Iran, including Bandar Anzali & Ramsar (per-humid), Gharakhil (sub-humid), Shiraz (semi-arid), Ahwaz (arid), and Yazd (extra-arid). The models’ performances were evaluated by the criteria percent bias (PB), root mean squared error (RMSE), normalized RMSE (NRMSE), and Nash-Sutcliff (NS) coefficient. Surveys confirm the high capability of the hybrid ANFIS-DE model in monthly ET0 forecasting; so that the DE algorithm was able to improve the accuracy of ANFIS, by 16% on average. Seasonal autoregressive integrated moving average (SARIMA) was the most suitable pattern among the time series stochastic models and superior to its competitors, ANFIS and ANFIS-DE. Consequently, the SARIMA was suggested more appropriate for monthly ET0 forecasting in all the climates, due to its simplicity and parsimony. Comparison between the different climates confirmed that the climate type significantly affects the forecasting accuracies: it’s revealed that all the models work better in extra-arid, arid and semi-arid climates, than the humid and per-humid areas.

A non-destructive dropped fruit impact signal imaging-based deep learning approach for smart sorting of kiwifruit

This paper focuses on developing an intelligent kiwifruit sorting system that uses an imaging method based on the dropped fruit impact signal, which combines with convolutional neural network (CNN) for the high-quality sorting of soft and hard kiwifruits. The kiwifruit sorting system detects the hardness of kiwifruit by analyzing the change in the force signal during the fall of kiwifruit on the sorting tray. The proposed signal imaging method constructs peak image (PI) by intelligently extracting the peak features of the data directly from the original time-domain force and force acceleration data in chronological order, and learns the features of PI for hardness classification by two-dimensional (2D) CNN. In addition, the performance of PI based on different peak, force signal or force acceleration signal in 2D CNN classifier is tested, and the 2D CNN classifier based on PI is compared with the 2D CNN classifier based on classical time-series imaging methods such as recurrence plot (RP) and Markov transition field (MTF), and further compared with the one-dimensional (1D) CNN model based on original time-domain features. A total of 18 CNN models are constructed for experiments in this paper, and the accuracy of the 2D CNN classifiers based on PI, MTF and RP are 98.89%, 94.75% and 98.34% at the time series length of 75 and the number of PI peaks of 5, respectively. The experimental results show that the PI-based 2D CNN classifier proposed in this paper achieves the best performances compared with other similar models, and demonstrates the feasibility of the MTF and RP-based 2D CNN classifier in kiwifruit hardness sorting.

Comparing forecasting accuracy of selected grey and time series models based on energy consumption in Brazil and India

The study compares the forecasting performance of the grey type models, represented by an optimized nonlinear grey Bernoulli model (ONGBM), a nonlinear grey Bernoulli model with particle swarm optimization (NGBM-PSO), and a standard GM with a classic time series model, represented by Auto-Regressive Integrated Moving Average (ARIMA). The models are compared based on simulations and energy consumption forecasting in Brazil and India at aggregate and disaggregate levels from 1992 to 2019. The study illustrates the picture of the Brazil and India energy consumption nexus at aggregate and disaggregated levels. The forecasting accuracy is compared using standard measures such as MAPE, MSE, RMSE, and normalized RMSE. The Diebold-Mariano test findings validates the ARIMA (1,1,1), GM (1,1), ONGBM (1,1) and NGBM (1,1)-PSO models' equal predictive performance. The models are used to compute forecast combinations, ensuring smaller forecasting errors than single models. Optimizing NGBM (1,1) using two algorithms ensures the highest forecasting efficiency for short time series. The results allow a recommendation to use ONGBM (1,1) and NGBM (1,1)-PSO models in the short-term forecasting of energy consumption and combine these forecasts with forecasts from GM (1,1) and ARIMA (1,1,1) models in practical applications.

Forecasting occurrence and quantity of monthly precipitation simultaneously while accounting for complex serial correlation

AbstractTweedie's compound Poisson regression models have been introduced in recent years to model and predict daily, monthly and seasonal precipitation data. Tweedie's compound Poisson regression analysis of precipitation data captures the relationship between the mean structure of precipitation and covariates while accounting for its right‐skewness and zero‐inflation appropriately, and thus extends trend analysis stage of classical time series techniques to handle right‐skewed time series data with possible zero‐inflation. A distinctive advantage of Tweedie's compound Poisson modelling of precipitation data is that the occurrence and quantity of precipitation can be simultaneously modelled using a single distribution; however, this approach ignores serial correlation between precipitation data observed over time. In this study, we propose to accommodate complex correlation structures of time series precipitation data using the Autoregressive Integrated Moving Average (ARIMA) models at the next stage as done in the classical time series analysis. Our analyses of monthly precipitation data in Australia demonstrated the usefulness of our two‐stage approach to prediction of precipitation.

Dynamics diagnosis of the COVID-19 deaths using the Pearson diagram

The pandemic COVID-19 brings with it the need for studies and tools to help those in charge make decisions. Working with classical time series methods such as ARIMA and SARIMA has shown promising results in the first studies of COVID-19. We advance in this branch by proposing a risk factor map induced by the well-known Pearson diagram based on multivariate kurtosis and skewness measures to analyze the dynamics of deaths from COVID-19. In particular, we combine bootstrap for time series with SARIMA modeling in a new paradigm to construct a map on which one can analyze the dynamics of a set of time series. The proposed map allows a risk analysis of multiple countries in the four different periods of the pandemic COVID-19 in 55 countries. Our empirical evidence suggests a direct relationship between the multivariate skewness and kurtosis. We observe that the multivariate kurtosis increase leads to the rise of the multivariate skewness. Our findings reveal that the countries with high risk from the behavior of the number of deaths tend to have pronounced skewness and kurtosis values.