Univariate Time Series Models Research Articles

Impact of unexpected conditions on construction cost forecasting performance: evidence from Europe

In order to prepare proper budgets, practitioners endeavor to forecast future cost fluctuations. In recent years, the COVID-19 pandemic and the Russia-Ukraine war have caused cost overruns in most construction projects, and there has been no effort to analyze the effects of such situations on cost forecasting performances. This paper aims to investigate the impact of unexpected conditions, such as pandemics and wars, on the construction cost forecasting performances of widely used univariate time series models. For this purpose, statistical forecasting models and two widely used machine learning techniques are employed for the construction cost forecasting of nine European economies in two scenarios: a stable period (2017–2020) and a fluctuating period (2020–2023). The results showed that the success of the models varied by region and period. Unexpected global conditions cause significant errors, especially in the medium and long term. There are significant increases in forecast errors between stable and fluctuating periods. This study provides comprehensive experimental results to determine the cost-forecasting risks of construction projects under unexpected global conditions. The investigation is limited to the European economies, which are mostly developed countries with low inflation rates. Therefore, the results should also be validated in underdeveloped and developing countries.

FORECASTING RAINFALL VOLUME IN SELANGOR WITH A COMBINED ARIMA MODEL

Flash flood is the most hazardous type of flooding, mainly caused by extensive rainfall. It also can cause significant harm to a community’s economy, ecology, and society without warning at an irrational pace. Therefore, this study was conducted to detect the time series element within the rainfall data, select the optimal model, and make predictions about the volume of rainfall in Selangor. A variety of univariate time series models were utilized, including the naïve model, decomposition model, Autoregressive Integrated Moving Average (ARIMA) model, exponential models, and combined models. Historical monthly rainfall data collected from Petaling station and Subang station from 2018 to 2022 were used to estimate the parameters of the models, and the model was evaluated for the smallest error of measurements. Previous research mostly focused on complex methodologies for forecasting rainfall. However, this research aimed to identify a simple tool for fast prediction of rainfall. The results showed that the combination of the ARIMA (2,0,3) model from Petaling Station and the ARIMA (4,0,4) model from Subang station were able to capture the trends and seasons in the time series with the lowest error of measurement on short-term predictions of rainfall volume. Furthermore, the study delves into the concept of combined time series models, which are blended using weighted performance measures to enhance prediction accuracy further. The research acknowledges certain limitations of univariate time series models, notably their inability to account for intricate interactions among environmental variables and potential long-term trends, such as those stemming from climate change. Overall, the study explores the potential of combining models to refine predictions for forecasting rainfall volume in Klang Valley.

Machine Learning and Deep Learning Data-Driven Residential Load Multi-Level Forecasting with Univariate and Multivariate Time Series Models Towards Sustainable Smart Homes

Machine Learning and Deep Learning Data-Driven Residential Load Multi-Level Forecasting with Univariate and Multivariate Time Series Models Towards Sustainable Smart Homes

Analysing the geographical impact on short-term solar irradiance forecasting using univariate and multivariate time series models

ABSTRACT Sustainable development is significantly hampered by air pollution brought on by greenhouse gas emissions. One of the primary contributors to air pollution is coal, which is used in thermal power plants. Substituting power plants from combustible sources to renewable energy sources can reduce air pollution to an extent. Solar energy is one of the most readily accessible renewable energy sources. High Global Horizontal Irradiance (GHI) is advantageous for deploying solar power plants, allowing maximum electricity production. However, the spatiotemporal factors affecting GHI are not analyzed, which can improve the forecasting accuracy. The proposed work analyzes the geographical impact of solar irradiation using statistical time-series models called ARIMA (Auto Regression Integrated Moving Average) and VAR (Vector Auto Regression) using four datasets from two different locations in India. The statistical time-series models can analyze the impact of the features and can show how the effect improves the performance of the multivariate model compared to the univariate model. The study aims to find the influence of the geographical factors of Global Horizontal Irradiance (GHI) based on the location chosen to build a solar power plant. The effectiveness of the univariate and multivariate models in forecasting solar irradiance is also discussed. The RMSE (root mean square error) analysis shows that the multivariate model VAR outperforms the univariate model in predicting solar irradiance. There is a 1–25% variance in the RMSE values between the ARIMA and VAR models at different locations and a 1–19% variance in MAE (Mean Absolute Error) value, while a 94–98% variance in MAPE (mean absolute percentage error) value. A deep analysis of factors that affect the precision of global horizontal irradiance (GHI) and the causal impact of variables is also analyzed using different variable impact tests. This variable selection is achieved by a feature selection algorithm called Maximum Relevance Minimum Redundancy (MRMR).

The modeling of earnings per share of Polish companies for the post-financial crisis period using random walk and ARIMA models

The proper forecasting of listed companies’ earnings is crucial for their appropriate pricing. This paper compares forecast errors of different univariate time-series models applied for the earnings per share (EPS) data for Polish companies from the period between the last financial crisis of 2008–2009 and the pandemic shock of 2020. The best model is the seasonal random walk (SRW) model across all quarters, which describes quite well the behavior of the Polish market compared to other analyzed models. Contrary to the findings regarding the US market, this time-series behavior is well described by the naive seasonal random walk model, whereas in the US the most adequate models are of a more sophisticated ARIMA type. Therefore, the paper demonstrates that conclusions drawn for the US might not hold for emerging economies because of the much simpler behavior of these markets that results in the absence of autoregressive and moving average parts.

Day-Ahead Electricity Demand Forecasting Using a Novel Decomposition Combination Method

In the present liberalized energy markets, electricity demand forecasting is critical for planning of generation capacity and required resources. An accurate and efficient electricity demand forecast can reduce the risk of power outages and excessive power generation. Avoiding blackouts is crucial for economic growth, and electricity is an essential energy source for industry. Considering these facts, this study presents a detailed analysis of the forecast of hourly electricity demand by comparing novel decomposition methods with several univariate and multivariate time series models. To that end, we use the three proposed decomposition methods to divide the electricity demand time series into the following subseries: a long-run linear trend, a seasonal trend, and a stochastic trend. Next, each subseries is forecast using all conceivable combinations of univariate and multivariate time series models. Finally, the multiple forecasting models are immediately integrated to provide a final one-day-ahead electricity demand forecast. The presented modeling and forecasting technique is implemented for the Nord Pool electricity market’s hourly electricity demand. Three accuracy indicators, a statistical test, and a graphical analysis are used to assess the performance of the proposed decomposition combination forecasting technique. Hence, the forecasting results demonstrate the efficiency and precision of the proposed decomposition combination forecasting technique. In addition, the final best combination model within the proposed forecasting framework is comparatively better than the best models proposed in the literature and standard benchmark models. Finally, we suggest that the decomposition combination forecasting approach developed in this study be employed to handle additional complicated power market forecasting challenges.

Factor price convergence in OECD economies: the case of three profit rates indicators

Abstract This paper examines profit rate (rates of return to capital) convergence in OECD economies in the periods of 1960-2016 and relevant sub-periods. It also performs comparison of profit rates convergence in selected developed and developing economies, using the data for 1973-2003 period. Economy-wide, productive economy and manufacturing rates are estimated and three convergence concepts are considered (beta, sigma, and stochastic convergence). We use a combination of cross-sectional and univariate time series models and density distribution analysis. For each profit rate measure, a strong evidence of beta convergence is provided. In contrast, sigma convergence is indicated only in the case of productive economy profit rate; while in other cases, sigma divergence or the absence of either convergence or divergence are likely. Stochastic convergence is present in a smaller number of economies and is confined to productive economy and manufacturing. The comparison of convergence dynamics in developed versus developing economies confirmed beta convergence in both groups and for their aggregate, but did not establish sigma convergence, given the significant diversity of economies and their different economic trajectories.

Bank crisis resolution in India

PurposeThe academic literature has examined why bank runs happen based on the work of 2022 Nobel Prize-winning economists Diamond and Dybvig. They have found the source of banking/financial crisis in terms of mismatch between liabilities (deposits being short term and savers wanting to short-term access to their money) and assets (long term and illiquid). The Lakshmi Vilas Bank (LVB) crisis intensified when it came under Prompt Corrective Action (PCA) of the Reserve Bank of India (RBI). This situation provides the opportunity to study whether the elements embodied in the theoretical models like Diamond and Dybvig hold true for LVB crisis. This study aims to examine the reasons for the demise of LVB in India using DuPont financial model, peer group analysis and time series structural break in crucial financial parameters.Design/methodology/approachThe study examines the reason for insolvency of LVB using financial ratios, financial models (DuPont), financial distress model (Z-score) and asset-liability management. The study also adopts univariate structural break models using quarterly financial data covering the key financial measures used in the RBI’s PCA framework.FindingsLVB crisis is like Diamond–Dybvig model, in the sense, savers requiring short-term access to their money (liquidity for their deposits) on the information of high non-performing assets, which further deteriorates the illiquid nature of loan portfolio (assets) of banks. The study finds its profit margin (net interest margin and non-interest margin) and managerial efficiency had started deteriorating since 2018. The study finds that LVB’s main weakness lies in its limited credit appraisal ability, its monitoring and weak internal controls. Lending to sensitive sectors (like real estate, capital markets and commodities) and exposure to large business groups also contributed to its weakness. The study also finds huge, elevated asset-liability mismatch, especially in the short-term maturity buckets. Using univariate econometric time series model, the study also confirms financial weakness being evident much earlier than the time when resolution was undertaken by the RBI through PCA.Research limitations/implicationsThe study has implications for analysing and monitoring financial distress of banks. The study also has implications for devising banking regulation and supervision.Originality/valueThe study brings in a perspective of the banking regulations using the application of PCA framework on a listed private sector bank. The authors combine an accounting ratio model and combine risk measures that could identify the incipient risks in a bank. The authors believe this will help in refinement of banking regulations and better monitoring mechanisms.

A New Family of Time Series to Model the Decreasing Relative Increment of Spreading of an Outbreak

Introduction: There are different mathematical models describing the propagation of an epidemic. These models can be divided into phenomenological, compartmental, deep learning, and individual-based methods. From other viewpoints, we can classify them into macroscopic or microscopic, stochastic or deterministic, homogeneous or heterogeneous, univariate or multivariate, parsimonious or complex, or forecasting or mechanistic. This paper defines a novel univariate bi-partite time series model able to describe spreading a communicable infection in a population in terms of the relative increment of the cumulative number of confirmed cases. The introduced model can describe different stages of the first wave of the outbreak of a communicable disease from the start to the end.
 Results: We use it to describe the propagation of various disease outbreaks, including the SARS (2003), the MERS (2018), the Ebola (2014-2016), the HIV/AIDS (1990-2018), the Cholera (2008-2009), and the COVID-19 epidemic in Iran, Italy, the UK, the USA, China and four of its provinces; Beijing, Guangdong, Shanghai, and Hubei (2020). In all mentioned cases, the model has an acceptable performance. In addition, we compare the goodness of this model with the ARIMA models by fitting the propagation of COVID-19 in Iran, Italy, the UK, and the USA.
 Conclusion: The introduced model is flexible enough to describe a broad range of epidemics. In comparison with ARIMA time series models, our model is more initiative and less complicated, it has fewer parameters, the estimation of its parameters is more straightforward, and its forecasts are narrower and more accurate. Due to its simplicity and accuracy, this model is a good tool for epidemiologists and biostatisticians to model the first wave of an epidemic.

Climate Change and its Impact on Sugarcane Production and Future Forecast in India: A Comparison Study of Univariate and Multivariate Time Series Models

Climate Change and its Impact on Sugarcane Production and Future Forecast in India: A Comparison Study of Univariate and Multivariate Time Series Models

On the Implementation of the Artificial Neural Network Approach for Forecasting Different Healthcare Events

The rising number of confirmed cases and deaths in Pakistan caused by the coronavirus have caused problems in all areas of the country, not just healthcare. For accurate policy making, it is very important to have accurate and efficient predictions of confirmed cases and death counts. In this article, we use a coronavirus dataset that includes the number of deaths, confirmed cases, and recovered cases to test an artificial neural network model and compare it to different univariate time series models. In contrast to the artificial neural network model, we consider five univariate time series models to predict confirmed cases, deaths count, and recovered cases. The considered models are applied to Pakistan's daily records of confirmed cases, deaths, and recovered cases from 10 March 2020 to 3 July 2020. Two statistical measures are considered to assess the performances of the models. In addition, a statistical test, namely, the Diebold and Mariano test, is implemented to check the accuracy of the mean errors. The results (mean error and statistical test) show that the artificial neural network model is better suited to predict death and recovered coronavirus cases. In addition, the moving average model outperforms all other confirmed case models, while the autoregressive moving average is the second-best model.

Multivariate time series models for mixed data

We introduce a general approach which unifies some previous attempts for modeling the dynamic of multivariate time series or for regression analysis when the data are of mixed type (binary/count/continuous). Our approach is quite flexible since conditionally on past values, each coordinate at time t can have a distribution compatible with a standard univariate time series model such as GARCH, ARMA, INGARCH or logistic models whereas past values of the other coordinates play the role of exogenous covariates in the dynamic. The simultaneous dependence in the multivariate time series can be modeled with a copula. Additional exogenous covariates are also allowed in the dynamic. We first study some usual stability properties of these models and then show that autoregressive parameters can be consistently estimated equation-by-equation using a pseudo-maximum likelihood method, leading to a fast implementation even when the number of time series is large. Moreover, we prove consistency results when a parametric copula model is fitted to the time series and in the case of Gaussian copulas, we show that the likelihood estimator of the correlation matrix is strongly consistent. We carefully check all our assumptions for two prototypical examples: a GARCH/INGARCH model and logistic/log-linear INGARCH model. Our results are illustrated with numerical experiments as well as two real data sets.

Multiple sclerosis incidence temporal trend in the Northeast of Iran: Using the Empirical Bayesian method.

In recent years dramatic changes in multiple sclerosis (MS) incidence have been reported in different provinces in Iran. This study was conducted to assess MS incidence temporal trends from March 21, 2005, to March 20, 2020, and provide a forecast until the end of 2025 in Shahroud county. This longitudinal study was carried out based on the data obtained from the MS registration system in Shahroud county. First, the annual incidence rates were calculated based on the year of diagnosis and smoothed using the Empirical Bayesian Method. Then temporal trends and annual percent change (APC) of MS incidence were analyzed using Joinpoint (JP) regression. Finally, the univariate time series model analysis was used to estimate the MS incidence trend until the end of 2025. A total of 234 newly diagnosed cases (60 [25.64%] males and 174 [74.36.4%] females) were examined in this study. The mean age of patients at the time of diagnosis was 31.40±3.78. It was 32.01±6.35 and 30.66±4.27 years for males and females, respectively (P<0.22). The mean annual MS incidence was 5.99±1.46, 3.03±0.21, and 8.98±2.79 per 100,000 in overall, males and females respectively. The MS incidence increased significantly from 5.67 (95% CI: 3.63-7.99) in 2005 to 7.58 (95% CI: 5.17-10.28) in 2020 with an APC of 4.5 (2.8 - 6.1). The MS incidence had a non-linear time trend in the study period and the best time trend fitted to the annual MS incidence trend was the non-linear quadratic curve. Based on this model, the annual MS incidence is expected to increase until the end of 2025. Shahroud county is one of the high-risk areas for MS and the increasing trend of MS incidence in it is similar to regional and global changes. This study, also, showed that MS incidence in Shahroud county will be increasing in the coming years.

Forecasting of methane gas in underground coal mines: univariate versus multivariate time series modeling

Forecasting of methane gas in underground coal mines: univariate versus multivariate time series modeling

Research of daily stock closing price prediction for new energy companies in China

&lt;abstract&gt;&lt;p&gt;Nowadays, China is developing new energy industries to reduce carbon emissions to meet the challenge of world climate change, so investors can consider to invest in stocks of Chinese new energy companies to gain income. In order to study how to forecast stock closing prices of new energy companies in China, we have chosen 12 representative companies, and first used autoregressive univariate time series models to predict the trends of the stock closing prices in the next month. The results show that Seasonal Autoregressive Integrated Moving Average model has the best out-of-sample trend prediction effect. Second, we use multivariate time series forecasting models to predict the stock closing prices of each day through external variables. The results show that Temporal Convolutional Attention Neural Networks has the best effect of out-of-sample prediction. We recommend that investors who are interested in investing in new energy companies in China first use the Seasonal Autoregressive Integrated Moving Average model to predict the short-term stock closing price trend in the future, and then use the Temporal Convolutional Attention Neural Networks model to predict the stock closing price on the next day to decide whether to invest.&lt;/p&gt;&lt;/abstract&gt;

Predicting daily emergency department visits using machine learning could increase accuracy

ObjectiveAdministrators and clinicians alike have attempted to predict emergency department visits for many years. The ability to predict or “forecast” ED visit volume can allow for more efficient resource allocation, including up-staffing or down-staffing, changing OR schedules, and predicting the need for significant resources. The goal of this study is to examine combinations of variables via machine learning to increase prediction accuracy and determine the factors that are most predictive of overall ED visits. As compared to a simple univariate time series model, we hypothesize that machine learning models will predict St. Joseph Mercy Ann Arbor's patient visit load for the emergency department (ED) with higher accuracy than a simple univariate time series model. MethodsUnivariate time series models for daily ED visits, including ARIMA, Exponential Smoothing (ETS), and Facebook Inc.'s prophet algorithm were estimated as a baseline comparison. Machine learning models, including random forests and gradient boosted machines (GBM), were trained using data from 2017 to 2018. After final models were created, they were applied to the 2019 data to determine how well these models predicted actual ED patient volumes in data not utilized during the model fitting process. The accuracy of the machine learning and time series models were assessed based on out-of-sample predictive accuracy, compared using root mean squared error (RMSE). ResultsUsing root mean squared error (RMSE) to assess out-of-sample predictive accuracy of the models, the results showed that the random forest model was the most accurate at predicting daily ED visits in the 2019 test set, followed by the GBM model. These performed only slightly better than the simple exponential smoothing model predictions. The ARIMA model performed poorly in comparison. The day of the week (likely capturing differences between weekdays and weekends) was found to be the most important predictor of patient volumes. Weather-related features such as maximum temperature and SFC pressure appeared to capture some of the seasonality trends related to changes in patient volumes. ConclusionsMachine learning models perform better at predicting daily patient volumes as compared to simple univariate time series models, though not by a substantial amount. Further research can help confirm these limited initial results. Gathering more training data and additional feature engineering could also be beneficial to training the models and potentially improving predictive accuracy.

Comparison study of transfer function and artificial neural network for cash flow analysis at Bank Rakyat Indonesia

&lt;span lang="EN-US"&gt;The cash flow analysis is essential to examine the economic flows in the financial system. In this paper, the financial dataset at Bank Rakyat Indonesia was used, it recorded the sources of cash inflow and outflow during a particular period. The univariate time series model like the autoregressive and integrated moving average is the common approach to build the prediction based on the historical dataset. However, it is not suitable to estimate the multivariate dataset and to predict the extreme cases consisting of nonlinear pairs between independent-dependent variables. In this study, the comparison of using two types of models i.e., transfer function and artificial neural network (ANN) were investigated. The transfer function model includes the coefficient of moving average (MA) and autoregressive (AR), which allows the multivariate analysis. Furthermore, the artificial neural network allows the learning paradigm to achieve optimal prediction. The financial dataset was divided into training (70%) and testing (30%) for two types of models. According to the result, the artificial neural network model provided better prediction with achieved root mean square error (RMSE) of 0.264897 and 0.2951116 for training and testing respectively.&lt;/span&gt;

Multi-Step Ahead Probabilistic Forecasting of Daily Streamflow Using Bayesian Deep Learning: A Multiple Case Study

In recent decades, natural calamities such as drought and flood have caused widespread economic and social damage. Climate change and rapid urbanization contribute to the occurrence of natural disasters. In addition, their destructive impact has been altered, posing significant challenges to the efficiency, equity, and sustainability of water resources allocation and management. Uncertainty estimation in hydrology is essential for water resources management. By quantifying the associated uncertainty of reliable hydrological forecasting, an efficient water resources management plan is obtained. Moreover, reliable forecasting provides significant future information to assist risk assessment. Currently, the majority of hydrological forecasts utilize deterministic approaches. Nevertheless, deterministic forecasting models cannot account for the intrinsic uncertainty of forecasted values. Using the Bayesian deep learning approach, this study developed a probabilistic forecasting model that covers the pertinent subproblem of univariate time series models for multi-step ahead daily streamflow forecasting to quantify epistemic and aleatory uncertainty. The new model implements Bayesian sampling in the Long short-term memory (LSTM) neural network by using variational inference to approximate the posterior distribution. The proposed method is verified with three case studies in the USA and three forecasting horizons. LSTM as a point forecasting neural network model and three probabilistic forecasting models, such as LSTM-BNN, BNN, and LSTM with Monte Carlo (MC) dropout (LSTM-MC), were applied for comparison with the proposed model. The results show that the proposed Bayesian long short-term memory (BLSTM) outperforms the other models in terms of forecasting reliability, sharpness, and overall performance. The results reveal that all probabilistic forecasting models outperformed the deterministic model with a lower RMSE value. Furthermore, the uncertainty estimation results show that BLSTM can handle data with higher variation and peak, particularly for long-term multi-step ahead streamflow forecasting, compared to other models.

Permanent layoff and consumer credit card loss forecasting

PurposeThe unemployment rate (UR) is the leading macroeconomic indicator used in the credit card loss forecasting. COVID-19 pandemic has caused an unprecedented level of volatility in the labor market variables, leading to new challenges to use UR in the credit risk modeling framework. This paper examines the dynamic relationship between the credit card charge-off rate and the unemployment rate over time.Design/methodology/approachThis study uses quarterly observations of charge-off rates on credit card loans of all commercial banks from Q1 1990 to Q4 2020. Univariate, multivariable, machine learning, and regime-switching time series modeling are employed in this research.FindingsThe authors decompose UR into two components – temporary and permanent UR. The authors find the spike in UR during COVID-19 is mainly attributed to the surge in temporary layoffs. More importantly, the authors find that the credit card charge-off rate is primarily driven by permanent UR while temporary UR has little predictive power. During recessions, permanent UR seems to be a stronger indicator than total UR. This research highlights the importance of using permanent UR for credit risk modeling.Originality/valueThe findings in the research can be applied to the credit card loss forecasting and CECL reserve models. In addition, this research also has implications for banks, macroeconomic data vendors, regulators, and policymakers.

Museums in the COVID-19 pandemic: the impact of economic stimulus policy and market strategy of the national palace museums in Taiwan

ABSTRACT This research examines the impact of the voucher-based economic stimulus policy and market strategy on museum traffic during the COVID-19 pandemic in Taiwan. We apply a univariate time-series model (ARMAX) to daily-visit data of the National Palace Museums in Taiwan during 2016–2021, showing results that the consumption voucher program effectively did boost the number of visitors amid overall declines in tourists due to the lockdown. A comparison among the National Palace Museum branches further reveals that holding more outdoor art events by one branch induced a more potent effect of the stimulus policy. The findings suggest that outdoor art events can be a helpful strategy for museums, which traditionally rely on indoor art exhibitions, to mitigate the impact of a pandemic.