Time Series Analysis Model Research Articles

Deployment of a business intelligence model to evaluate Iranian national higher education

Abstract Higher education plays an important role in the political and socio-economic development of countries. Developing countries experience many significant challenges when it comes to national higher education programs; issues such as financial insecurity, poor managerial practices, and system inefficiencies are some of the obstacles that developing countries have yet to overcome. Resource allocation, technical efficiency, and managerial effectiveness are some of the significant objectives of government national higher education programs for developing countries-including those in the Middle East. The distribution of relevant data sources and the complexity of dynamism in higher education systems allows for an integrated intelligent system with a multi-dimensional view of the current situation to be built. This study proposes a business intelligence-based model to support the monitoring of higher education indicators and enable the forecasting of future trends through the integration of heterogeneous internal and external data sources. In the case study on Iranian higher education indicators, a prototype system was designed and implemented to evaluate the proposed model and its efficiency in practice. After monitoring the indicators using online analytical processing, several indicators were used to forecast trends by time series analysis models. The developed system attempts to provide an integrated view of the Iranian higher education system in comparison with other neighboring countries. The results emphasize that while higher education in Iran, particularly in the area of science and engineering, is a benchmark in the scientific community, the intense level of brain drain is increasing at an alarming rate.

Time Series Data Analysis for Stock Market Prediction

Prediction of the stock market is gathering of previous historical data of shares or any traded securities in the market and trying to forecast the future value of the company's share/securities. This stock prediction using historical time series data, based on certain preconceived notions and some constant factors and hence becomes difficult to predict for long term range. (Figlewski, 1994) (Mondal, 2014).Time series analysis like moving average, autoregression integrated. Moving Average and Holtwinters Models are being used. This paper proposes the best fit model for stock prediction using BSE (Bombay Stock Exchange data ) for 10 years. Purpose: A typical mistake investor does many a time a wrong timing of investment. Investing at an appropriate time can get gains. Investors often fail to understand when to buy the stock, when to sell, how long to hold as many times misconceptions. (Grinblatt, 2000). Thus we have attempted to find the best method for Prediction. Also attempted to predict nearing to accurate Prediction and can generate profits to the investors. (Zweig, 1973).However, the limitation is the Time series analysis model is helpful in short period forecasting (Alwadi, 2018, October). (Chng, 2019) (Ariyo, 2014, March). Stock market prediction for the more extended period is a challenge because of the very nature of Data without seasonality or any trend. Significance of the Study This paper proposes the best fit model for stock prediction using BSE (Bombay Stock Exchange data ) share data for 10 years. The results can be useful to the stock analyst, Government, investor, management and all the stakeholders. However, since stock market data do not give uniform patterns, different analysis is tested to predict the future value and get reasonably accurate results. Objective of the study: The objective of this research is to find out pattern from 10 years BSE data and estimating future price. This paper thus focuses on predicting an approximate value of share price and also best fit method for Prediction

A Statistical Time-Frequency Model for Non-stationary Time Series Analysis

Time-frequency analysis (TFA) plays an important role in various engineering and biomedical fields. For a non-stationary time series, a common practice is to divide data into segments under the piecewise stationarity assumption and perform TFA for each segment. In this article, we propose a three-layer latent variable model that relaxes such an assumption and therefore provides a more flexible solution to identify the frequency components and characterize their evolution over time for non-stationary time series with multi-component signals. Our proposed model is built upon hierarchical Dirichlet process (HDP), hidden Markov model (HMM) and extended time-varying autoregressive (ETVAR) model. The proposed approach does not impose any restrictions on the number and locations of segments, or the number and values of the frequency components within a segment. Both the simulation studies and real data applications demonstrate the superiority of the proposed method over existing methods.

TIME SERIES MODELS FOR FORECASTING EXCHANGE RATES

This paper investigates the behavior of daily exchange rate of the Georgian Currency LARI (GEL) exchange rate against the USDand EUR. To forecast exchange rates there are numerous models, which tend from very simple to very complicated models for analysis of GEL/USD and GEL/EUR time series variable. The objective of this paper is to com- pare the performance of individual time series models for predictingexchange rates. We will investigate the application of following time series analysis models: moving average, ex- ponential smoothing, double exponential smoothing adjust- ed for trend, time-series decomposition models, and ARIMA class models. The forecasting ability of these models is subsequently assessed using the symmetric loss functions which are the Mean Absolute Percentage Error (MAPE), the Mean Absolute deviation (MAD), and the Mean Squared error /deviation (MSE/MSD). In some cases, predicting the direction of exchange rate change may be valuable and profitable. Hence, it is reasonable to look at the frequency of the correctpredicted direction of change by used models, for short - FCPCD. An exchange rate represents the price of one currency in terms of another. It reflects the ratio at which one currency can be exchanged with another currency. Exchange rates forecasting is a very important and challenging subject of finance market, to determine optimal government policies as well as to make business decisions. This is important for all that firms which having their business spread over different countries or for that which raise funds in different currency. Business people mainly use exchange rates forecasting results in following types of decisions like choice currency for invoicing, pricing transactions, borrowing and landing currency choice, and management of open currency positions. The forex market is made up of banks, commercial companies, central banks, investment management firms, hedge funds, and retail forex brokers and investors. Forecasting the short- run fluctuations and direction of change of the currency ex- change rates is important for all these participates. The main goal of this study is to forecast of future ex- change rate trends by using currency rates time-series, rep- resenting past trends, patterns and waves. The monetary policy of the National Bank of Georgia since 2009 have been followed the inflation targeting regime, where exchange rate regime is floating - change of exchange rate is free. The offi- cial exchange rate of the Georgian GEL against the USD is cal- culated each business day. The official exchange rate of GEL against USD is calculated as the average weighted exchange rate of the registered spot trades on the interbank market functioning within the Bloomberg trade platform. Then, the official exchange rate of GEL against other foreign currencies is determined according to the rate on international markets on the basis of cross-currency exchange rates.

Multiresolution mode decomposition for adaptive time series analysis

This paper proposes the multiresolution mode decomposition (MMD) as a novel model for adaptive time series analysis. The main conceptual innovation is the introduction of the multiresolution intrinsic mode function (MIMF) of the form∑n=−N/2N/2−1ancos⁡(2πnϕ(t))scn(2πNϕ(t))+∑n=−N/2N/2−1bnsin⁡(2πnϕ(t))ssn(2πNϕ(t)) to model nonlinear and non-stationary data with time-dependent amplitudes, frequencies, and waveforms. The multiresolution expansion coefficients {an}, {bn}, and the shape function series {scn(t)} and {ssn(t)} provide innovative features for adaptive time series analysis. For complex signals that are a superposition of several MIMFs with well-differentiated phase functions ϕ(t), a new recursive scheme based on Gauss-Seidel iteration and diffeomorphisms is proposed to identify these MIMFs, their multiresolution expansion coefficients, and shape function series. Numerical examples from synthetic data and natural phenomena are given to demonstrate the power of this new method.

Effect of Monitoring the Compliance of Banning Alcohol Sales to Minors in the Volume of Underage Acute Alcohol Intoxication Cases in Mexico: A Controlled ITSA Analysis

In Mexico, underage alcohol consumption is a major public health problem. Notwithstanding the complex character of alcohol consumption, evidence indicates that enforcing the policies aimed to control the availability of alcohol has produced positive outcomes and has proved to be one of the most cost-effective strategies for curbing consumption, as well as they have succeeded in reducing health and social problems associated to alcohol use. Evaluate the effect of enforcing compliance with minimum-age restrictions for the sale of alcoholic beverages in the number of cases of acute alcohol intoxication in underage population in Mexico. A controlled interrupted time series analysis (ITSA) model was resorted in order to assess the effect of enforcing the ban of alcohol sales to underage individuals at national level in Mexico. The outcome variable was cases of acute alcohol intoxication (F10.0) discharged in emergency services in the group from 10 to 19 years of age. We found a statistically significant drop (16.4%) in the cases of acute alcohol intoxication immediately after the program implementation. This result suggests that monitoring and enforcing strategies to ban alcohol sales to underage individuals can provide an effective tool to reduce the negative outcomes of underage alcohol consumption.

Impact of direct-to-consumer drug advertising during the Super Bowl on drug utilization.

Direct-to-consumer advertising (DTCA) of prescription drugs impacts patients' requests for medications, and clinician prescribing. However, the impact of DTCA during the Super Bowl has not been previously described. Evaluate the impact of prescription drug DTCA during the Super Bowl on drug utilization using 2014-2016 Medicare data. Efinaconazole was advertised during Super Bowls XLIX (02/01/2015) and L (02/07/2016). The number of prescriptions for efinaconazole and for a comparator drug, tavaborole, were calculated in 31-day intervals from July 2014-December 2016. Interrupted time-series analysis models were created to test changes in trends of prescriptions for efinaconazole and tavaborole. Following Super Bowl XLIX, the number of prescriptions per 100,000 Medicare beneficiaries increased by 91% for efinaconazole, and 275% for tavaborole. After Super Bowl L, the number of prescriptions increased significantly for efinaconazole (p-value<0.001), but not for tavaborole (p=0.70). Interrupted time-series analyses estimated that, in the absence of DTCA during Super Bowl XLIX, prescriptions for efinaconazole would have increased by 40%, instead of the observed 91%. For tavaborole, prescriptions would have increased by 90% instead of 275%. DTCA during the Super Bowl resulted in sharp increases in utilization of the prescription drug advertised, which supports further regulation of DTCA.

Trends of non-communicable diseases and public health concerns of the people of northeastern Nigeria amidst the Boko Haram insurgency

The ongoing Boko Haram insurgency in northeastern Nigeria has depleted the country’s capability to deliver quality healthcare to her citizenry. The ailing health sector is overwhelmed with a rising incidence and prevalence of infectious and non-communicable diseases. The aim of this paper was to determine the trend of kidney disease, end-stage renal disease (ESRD), anaemia, malnutrition and human immunodeficiency virus (HIV) in northeastern Nigeria. Data covering different periods between 1999 and 2017 were obtained from the University of Maiduguri Teaching Hospital (UMTH), Maiduguri in northeastern Nigeria. This hospital is the largest tertiary healthcare facility in the region. The data were demographic in nature, i.e. sex, age or simply the registered cases. A quadratic model of time-series analysis was used to create the various trends of the respective diseases with the aid of Minitab software (version 18.0). t-Tests and analysis of variance (ANOVA) were also performed, with a p-value of less than or equal to 0.05 being considered significant. About three out of every five patients treated for kidney disease were male, while three out of every five patients treated for HIV were female. Patients between the ages of 31 and 50 years were at the greatest risk of developing kidney disease. The mean distribution of disease incidence was the same for both sexes for kidney disease, anaemia, malnutrition and HIV. The mean distribution disease incidence was different between age groups for kidney disease but the same for anaemia. The incidence of anaemia and malnutrition reported was almost the same for both sexes. Children under the age of 10 years old were at the most risk of anaemia, with the distribution decreasing with increasing age. The trend analysis of the incidence of kidney disease, ESRD, anaemia and malnutrition showed that these ailments were on the increase, while the number of patients that were on antiretroviral therapy (ART) was on the decline in northeastern Nigeria. These increasing trends are evidences of the effect of Boko Haram on the public health of the people of northeastern Nigeria and will continue to be a public health concern for the region and the country as a whole.

PC2A: Predicting Collective Contextual Anomalies via LSTM With Deep Generative Model

Proactive anomaly detection and diagnosis play an essential role in ensuring the security and stability of a large-scale information technology (IT) system with thousands or even millions of components that are interacting with each other. Collective contextual anomalies (CCAs) carry the characteristics of both collective and contextual anomalies. This type of anomalies is common in IT system monitoring, often manifested as security risks to be ameliorated, service outages to be eliminated, or stragglers to be mitigated. However, most existing studies emphasize primarily on the detection of point anomalies while the prediction or early detection of CCA has been an underexplored topic. In this paper, we propose a framework for discovering and studying CCAs in multiple time series based on a combination of semi-supervised deep learning, time series modeling, and graph analysis. A primary advantage of the proposed framework is that it can effectively predict CCAs with no human intervention. In addition, the performance of the proposed method can be further enhanced via learning from a small amount of anomalous training data, if it is available. Finally, the proposed framework is of low computational complexity and is thus particularly suitable for large-scale data streams. Simulation studies are carried out to reveal the superior performance of the proposed method and underscore the significant benefits of combining deep neural networks with time series analysis and graph models for the prediction and analysis of CCAs.

Climate modeling using vector moving average autoregressive

Information on weather and climate is a major need to support the smooth running of activities in various sectors, one of which is the agricultural sector. The agricultural sector is very dependent on climate change, where extreme climate change can cause storms, floods, droughts, resulting in crop failure. Therefore, climate information is very important, with information on climate change predictions, it can be used to determine when the right planting time is so that the quality and yield of agricultural production increases. The purpose of this study is to model climate data rainfall, humidity, maximum temperature, maximum temperature, and the length of solar radiation using a Vector Autoregressive Integrated Moving Average (VARIMA) approach. VARIMA is a time series model involving multivariate time series data, where the VARIMA model is a vector form of the ARIMA model. The time series analysis model requires data to meet stationary requirements. The results of the analysis obtained stationary climate data after being differed once, so the differencing data results that will be used in modelling. The time series climate data model was combined until the fourth lag based on SAS software assistance and the results were obtained namely the VARIMA model (1,1,1) with MSE values of 6045,33 and AICC namely 17,97823 and VARIMA (2,1,0) with MSE values namely 7432,434 and AICC which is 17,978023. Based on the two models, one of the best models was chosen with the smallest MSE and AIC value criteria. Thus the best VARIMA model that can be used to predict climate data in Merauke is the VARIMA (1,1,1) model.

Application of Exponential Smoothing Models and Arima Models in Time Series Analysis from Telco Area

Application of Exponential Smoothing Models and Arima Models in Time Series Analysis from Telco Area

Time Series Analysis of Interest Rates Volatility and Stock Returns in Ghana

The study utilized time series analysis models and employed the Johansen’s cointegration procedure and the vector error correction model to examine the short-run and long-run dynamics of the relationship between interest rates and stock market returns. The results of this study show that contrary to popular evidence from extant research, interest rate changes positively and significantly affect stock market returns in the long run and the deviation from the long-run equilibrium is corrected each period following a shock to the stock market in the short run. The positive linkages between interest rate changes and stock market outturns may be explained by the relative strength of banking stocks on the Ghana Stock Exchange. The analysis shows that as the long-run equilibrium is approached, the deviations in the short term decrease significantly.

Applying Auto-Regressive Model's Yule-Walker Approach to Amyotrophic Lateral Sclerosis (ALS) patients' Data.

The purpose of this study is to identifying time series analysis and mathematical model fitting on electroencephalography channels that are placed on Amyotrophic Lateral Sclerosis (ALS) patients with P300 based brain-computer interface (BCI). Amyotrophic Lateral Sclerosis (ALS) or motor neuron diseases are a rapidly progressing neurological disease that attacks and kills neurons responsible for controlling voluntary muscles. There is no cure and treatment effective to reverse, to halt the disease progression. A Brain- Computer Interface enables disable person to communicate & interact with each other and with the environment. To bypass the important motor difficulties present in ALS patient, BCI is useful. An input for BCI system is patient's brain signals and these signals are converted into external operations, for brain signals detection, Electroencephalography (EEG) is normally used. P300 based BCI is used to record the reading of EEG brain signals with the help of non-invasive placement of channels. In EEG, channel analysis Autoregressive (AR) model is a widely used. In the present study, Yule-Walker approach of AR model has been used for channel data fitting. Model fitting as a form of digitization is majorly required for good understanding of the dataset, and also for data prediction. Fourth order of the mathematical curve for this dataset is selected. The reason is the high accuracy obtained in the 4th order of Autoregressive model (97.51±0.64). In proposed Auto Regressive (AR) model has been used for Amyotrophic Lateral Sclerosis (ALS) patient data analysis. The 4th order of Yule Walker auto-regressive model is giving best fitting on this problem.

Detection of Liquefaction Phenomena from the 2017 Pohang (Korea) Earthquake Using Remote Sensing Data

On 15 November 2017, liquefaction phenomena were observed around the epicenter after a 5.4 magnitude earthquake occurred in Pohang in southeast Korea. In this study, we attempted to detect areas of sudden water content increase by using SAR (synthetic aperture radar) and optical satellite images. We analyzed coherence changes using Sentinel-1 SAR coseismic image pairs and analyzed NDWI (normalized difference water index) changes using Landsat 8 and Sentinel-2 optical satellite images from before and after the earthquake. Coherence analysis showed no liquefaction-induced surface changes. The NDWI time series analysis models using Landsat 8 and Sentinel-2 optical images confirmed liquefaction phenomena close to the epicenter but could not detect liquefaction phenomena far from the epicenter. We proposed and evaluated the TDLI (temporal difference liquefaction index), which uses only one SWIR (short-wave infrared) band at 2200 nm, which is sensitive to soil moisture content. The Sentinel-2 TDLI was most consistent with field observations where sand blow from liquefaction was confirmed. We found that Sentinel-2, with its relatively shorter revisit period compared to that of Landsat 8 (5 days vs. 16 days), was more effective for detecting traces of short-lived liquefaction phenomena on the surface. The Sentinel-2 TDLI could help facilitate rapid investigations and responses to liquefaction damage.

Changes in the Management of Children With Brief Resolved Unexplained Events (BRUEs).

In May 2016, the American Academy of Pediatrics published a clinical practice guideline for brief resolved unexplained events (BRUEs). We evaluated for changes in the management of BRUE after guideline publication. Using a pediatric multicenter administrative database, we compared rates of admission, testing, revisits, and diagnoses in patients diagnosed with a BRUE or apparent life-threatening event (ALTE) during 2017 with rates of admission, testing, revisits, and diagnoses in patients diagnosed with ALTE during 2015. We used interrupted time series analysis to test if the guideline was associated with changes in admission rate for all patients with ALTE or BRUE between 2015 and 2017. We stratified analyses by age (0-60 and 61-365 days). A total of 9501 patients were included (5608 in 2015 and 3893 in 2017). The admission rate decreased by 5.7% (95% confidence interval, 3.8% to 7.5%) for infants 0 to 60 days and by 18.0% (95% confidence interval, 15.3% to 20.7%) for infants 61 to 365 days from 2015 to 2017. Patients in 2017 had lower rates of EEG, brain MRI, chest radiography, laboratory testing, and urinalyses compared with patients in 2015. In the interrupted time series analysis model (n = 13 977), guideline publication was associated with decreasing admission rates (0.2% per week) for infants 61 to 365 days (P < .001). Compared with patients evaluated in 2015, patients with BRUE or ALTE in 2017 have lower rate of admissions and testing. Findings may be due to changes in the definition of BRUE and guideline recommendations.

Time-Series-Based Model and Validation for Prediction of Exposure to Wideband Radio Frequency Electromagnetic Radiation

Strength of wideband electromagnetic radiations (EMRs) generated by telecommunication systems is a time-varying process typically characterized by a seasonal behavior due to daily and weekly variations of traffic in mobile networks summed to a stationary background generated by constant-power broadcasting systems. Besides a few papers showing the existence of peak and off-peak hours in the daily exposure, there is a significant lack of scientific literature investigating how exposure changes over time, although this is key to the three main objectives: 1) to determine the time intervals when exposure levels are expected to be the highest, 2) to assess worst case exposure and ensure that measurements are run in a conservative scenario to comply with limits enforced by national laws and European guidelines, and 3) to highlight variations in traffic to allocate network resources dynamically to optimize performance. We present a time-series analysis model that makes use of measured data to build a predictor of the exposure to wideband EMR. The main result is that exposure can be predicted by a model composed of three components capturing the average exposure, the daily and half-daily periodicity (seasonal behavior), and the correlations among residuals [modeled by an AR(2) for weekdays and Sunday, and by an AR(1) for Saturday, respectively]. We observed the strength of the periodic components differs more on weekdays compared to weekends. The forecasting model can be particularly useful to predict when exposure is expected to be the highest during the day, so that measurements carried out at that time will be conservative with respect to measurements run at other times of the day, and will therefore serve as the worst case for the assessment of compliance with exposure limits.

Predictive models for wastewater flow forecasting based on time series analysis and artificial neural network.

Wastewater flow forecasting is key for proper management of wastewater treatment plants (WWTPs). However, to predict the amount of incoming wastewater in WWTPs, wastewater engineers face challenges arising from numerous complexities and uncertainties, such as the nonlinear precipitation-runoff relationships in combined sewer systems, unpredictability due to aging infrastructure, and frequently inconsistent data quality. To address such challenges, a time series analysis model (i.e., the autoregressive integrated moving average, ARIMA) and an artificial neural network model (i.e., the multilayer perceptron neural network, MLPNN) were developed for predicting wastewater inflow. A case study of the Barrie Wastewater Treatment Facility in Barrie, Canada, was carried out to demonstrate the performance of the proposed models. Fifteen-minute flow data over a period of 1 year were collected, and the resampled daily flow data were used to train and validate the developed models. The model performances were examined using root mean square error, mean absolute percentage error, coefficient of determination, and Nash-Sutcliffe efficiency. The results indicate that both models provided reliable forecasts, while ARIMA showed a slightly better performance than MLPNN in this case study. The proposed models can provide useful decision support for the optimization and management of WWTPs.

Time Series Prediction Algorithm for Intelligent Predictive Maintenance

Predictive maintenance aims to find out when the target device (TD) is in the sick state and almost entering the dead state before its actual occurrence to conduct just-in-time maintenance, so as to avoid unexpected TD down time. In this way, not only tool availability and manufacturing quality are improved, but the additional cost of excessive maintenance in preventive maintenance strategy can also be reduced. Among the predictive maintenance technologies proposed by many scholars, exponential model was commonly applied to predict the remaining useful life (RUL) of TD. However, due to the algorithm limitations, when TD is about to die, whether the TD's aging feature suddenly rises or becomes smooth, the exponential model may not be able to keep up with the real-time prediction or even falsely predicts long RUL. To solve the problem of inaccurate RUL prediction, the authors propose the time series prediction (TSP) algorithm. TSP applies the time series analysis model built by information criterion to adapt to the complicated future trend of solving TD fault prediction. Also, the Pre-Alarm Module (PreAM) to make alert of immediate maintenance when a TD is likely to shut down shortly as well as the Death Correlation Index (DCI) to reveal the possibility of entering the dead state are proposed in this work. How to select the most effective predictors and adjust the predictor weights to construct high-performance prediction model are also illustrated in this letter with the tools in various industries (such as solar-cell manufacturing and machine tool industry) being the examples of the TSP algorithm.

Trends and prediction of antimicrobial susceptibility in urinary bacteria isolated in European emergency departments: the EuroUTI 2010-2016 Study.

To assess recent trends in susceptibility to antibiotics among urinary isolates isolated in European emergency departments (EDs) and to identify isolates with a high (90% or more) predicted probability of susceptibility to fluoroquinolones or third-generation cephalosporins (3GCs). In this cross-sectional study, we included urine cultures obtained from adult patients between 2010 and 2016 in 24 European EDs. Temporal trends were assessed using time-series analysis and multivariate logistic models. Multivariate logistic models were also used to predict susceptibility to fluoroquinolones or 3GCs from patient age and sex, year, month and ED. We included 88242 isolates. Time-series analysis found a significant increase in susceptibility to fluoroquinolones and no significant trend for susceptibility to 3GCs. Adjusting for patient age and sex, ED and organism, multivariate models showed that susceptibility to 3GCs decreased from 2014 to 2016, while susceptibility to fluoroquinolones increased in 2015 and 2016. Among isolates from 2016, multivariate models predicted high probability of susceptibility to fluoroquinolones in 11% of isolates (positive predictive value 91%) and a high probability of susceptibility to 3GCs in 35% of isolates (positive predictive value 94%). Susceptibility of ED urinary isolates to fluoroquinolones increased from 2014, while susceptibility to 3GCs decreased from 2015. Predictive models identified isolates with a high probability of susceptibility to fluoroquinolones or 3GCs. The ability of such models to guide the empirical treatment of pyelonephritis in the ED remains to be determined.

The its.analysis R Package – Modelling Short Time Series Data

Modelling data over time requires a set of unique and bespoke processes in order to deal with problems induced by temporal dependency and autocorrelation. Methods are well developed and widely applied for working with such data when sample sizes are high, but the issue of how to fit models which are methodologically robust but not under-powered or over-fitted on short time series data remains vexing. This research proposes an interrupted time series analysis model solution to this problem, and uses a Type II Sum Squares ANCOVA Lagged Dependent Variable, variance-centric approach as part of a newly introduced R package - its.analysis. Using this model switches the null hypothesis situation to a much more reliable test and allows for a much more flexible approach to adding covariates in small sample conditions. The model performs very well under test conditions, appears more conservative than existing alternative techniques, and as such is recommended to researchers for future analysis of temporal data where observations are limited (between 15 and 45 observations).