Beyond the beaten paths of forecasting call center arrivals: on the use of dynamic harmonic regression with predictor variables

Modern call centers require precise forecasts of call and e-mail arrivals to optimize staffing decisions and to ensure high customer satisfaction through short waiting times and the availability of qualified agents. In the dynamic environment of multi-channel customer contact, organizational decision-makers often rely on robust but simplistic forecasting methods. Although forecasting literature indicates that incorporating additional information into time series predictions adds value by improving model performance, extant research in the call center domain barely considers the potential of sophisticated multivariate models. Hence, with an extended dynamic harmonic regression (DHR) approach, this study proposes a new reliable method for call center arrivals’ forecasting that is able to capture the dynamics of a time series and to include contextual information in form of predictor variables. The study evaluates the predictive potential of the approach on the call and e-mail arrival series of a leading German online retailer comprising 174 weeks of data. The analysis involves time series cross-validation with an expanding rolling window over 52 weeks and comprises established time series as well as machine learning models as benchmarks. The multivariate DHR model outperforms the compared models with regard to forecast accuracy for a broad spectrum of lead times. This study further gives contextual insights into the selection and optimal implementation of marketing-relevant predictor variables such as catalog releases, mail as well as postal reminders, or billing cycles.

Community factors associated with local epidemic timing of respiratory syncytial virus: A spatiotemporal modeling study

Respiratory syncytial virus (RSV) causes a large burden of morbidity in young children and the elderly. Spatial variability in the timing of RSV epidemics provides an opportunity to probe the factors driving its transmission, including factors that influence epidemic seeding and growth rates. Using hospitalization data from Connecticut, New Jersey, and New York, we estimated epidemic timing at the ZIP code level using harmonic regression and then used a Bayesian meta-regression model to evaluate correlates of epidemic timing. Earlier epidemics were associated with larger household size and greater population density. Nearby localities had similar epidemic timing. Our results suggest that RSV epidemics grow faster in areas with more local contact opportunities, and that epidemic spread follows a spatial diffusion process based on geographic proximity. Our findings can inform the timing of delivery of RSV extended half-life prophylaxis and maternal vaccines and guide future studies on the transmission dynamics of RSV.

016 Associations between Circadian Melatonin and Temperature Amplitudes during Constant Routine

Introduction Circadian amplitude measures the strength or robustness of a rhythm and changes in amplitude may have implications for health. Large individual differences in melatonin amplitude are recognized. Here we aimed to determine the strength of relationships between melatonin and the core body (CBT) and distal-proximal skin temperature gradient (DPG) amplitudes during a constant routine protocol. Additionally, we determined the best fitting harmonic model for the DPG circadian rhythm. Methods 17 young healthy adults [13 males (22.3±3.9yr;mean±SD)] completed a 28-hr constant routine protocol after maintaining 8h habitual sleep schedules for one week at home. Endogenous circadian amplitudes of melatonin and CBT were fit with standard three- and dual-harmonic linear regression models, respectively. The DPG amplitude was analyzed with both dual and three-harmonic regression models to determine which model produced the best fit. Results The DPG was best fit by a three-harmonic regression model with significantly lower standard deviation and higher signal-to-noise ratio compared to the 2-harmonic model (both p<0.05) as well as by visualization of the fitted curves. Melatonin, CBT and DPG amplitudes were not found to be associated with each other during constant routine (all r<0.37; all p>0.10). Conclusion While it is common for melatonin and body temperature circadian phase estimates to be used interchangeably, non-significant findings for associations between circadian amplitudes of melatonin, CBT and DPG indicate that these markers may not provide similar information about circadian amplitude. Further, research is needed to explore possible associations between individual differences in melatonin, CBT and DPG amplitudes with other physiological and behavioral outcomes to determine which measure(s) of circadian amplitude may be functionally relevant. Support (if any) NIH R01 HL081761

A dynamic harmonic regression approach for bridge structural health monitoring

Structural damage in a bridge is defined as a significant deviation in the structural response from its standard operating conditions, not explainable by variations in external environmental and operational effects. However, environmental effects such as temperature fluctuations can cause significant seasonal variations in the structural response of a bridge and can mask its changes due to structural damage. This poses a challenge for structural health monitoring of bridges where reliable diagnosis of damage or deterioration is often related to isolation of the responses. To address it, a statistical damage-detection methodology is introduced where strain data are modelled using a dynamic harmonic regression time-series model. Prediction intervals of multi-step ahead forecasts from the dynamic harmonic regression model are then used as statistical control limits within which the observed phenomenon should fall under standard operating conditions. This single recursive structural health monitoring framework for automatic fitting and multi-step ahead forecasting of 1-min interval time-series strain data includes recorded temperature values and diurnal trends as regressors in the model to account for environmental variations. The potential of this method as a robust automatic structural health monitoring strategy is demonstrated on strain data sampled at 1-min interval from a full-scale damaged pre-stressed concrete bridge – before, during and after repair. The proposed method can capture both sudden and daily changes in structural response due to temperature effects, and a rolling multi-step ahead interval forecast was able to identify damage on back-cast data transitioning from a healthy state to a damaged state.

Modeling of the Weekly Variation of the Reported COVID-19 Cases as A Potential Indicator of the Surveillance System Accuracy

Periodical daily variation in the number of reported COVID-19 cases within weeks is a common observation in global and national statistics. This variation may imply that the day of week has a significant role in the number of reported cases. We compared the pattern in some countries with an acceptable surveillance system. Data of 18 European and North American countries between 6 Mar and 8 Nov 2020 were extracts. Harmonic regression models were used to quantify the peak day, the absolute intensity and the average of coefficient of variation within weeks (ACVW) classified by country. In eight countries, the within week variation was statistically significant, the maximum and minimum number reported cases were in Saturday and Monday respectively, however, this pattern varied among countries. The maximum of ACVW was observed in Belgium and France, while it was minimum in Russia. The level of intensity of infection had a positive association with the ACVW (r = 0.54, p-value = 0.021). The observed variation and its pattern may show that the coverage or the tidiness of COVID-19 surveillance system fluctuates in different days of week. In addition, we suggest that the level of this fluctuation might be used as an accuracy indicator of the surveillance system.

Impact of Seasonality and Influenza Rates on Interventions to Reduce Hospital-Acquired Clostridioides difficile Rates

Background: Hospital-acquired Clostridioides difficile infection (HA-CDI) rates are highly variable over time, posing problems for research assessing interventions that might improve rates. By understanding seasonality in HA-CDI rates and the impacts that other factors such as influenza admissions might have on these rates, we can account for them when establishing the relationship between interventions and infection rates. We assessed whether there were seasonal trends in HA-CDI and whether they could be accounted for by influenza rates. Methods: We assessed HA-CDI rates per 10,000 patient days, and the rate of hospitalized patients with influenza per 1,000 admissions in 4 acute-care facilities (n = 2,490 beds) in Calgary, Alberta, from January 2016 to December 2018. We used 4 statistical approaches in R (version 3.5.1 software): (1) autoregressive integrated moving average (ARIMA) to assess dependencies and trends in each of the monthly HA-CDI and influenza series; (2) cross correlation to assess dependencies between the HA-CDI and influenza series lagged over time; (3) Poisson harmonic regression models (with sine and cosine components) to assess the seasonality of the rates; and (4) Poisson regression to determine whether influenza rates accounted for seasonality in the HA-CDI rates. Results: Conventional ARIMA approaches did not detect seasonality in the HA-CDI rates, but we found strong seasonality in the influenza rates. A cross-correlation analysis revealed evidence of correlation between the series at a lag of zero (R = 0.41; 95% CI, 0.10–0.65) and provided an indication of a seasonal relationship between the series (Fig. 1). Poisson regression suggested that influenza rates predicted CDI rates (P < .01). Using harmonic regression, there was evidence of seasonality in HA-CDI rates (2 [2 df] = 6.62; P < .05) and influenza rates (2 [2 df] = 1,796.6; P < .001). In a Poisson model of HA-CDI rates with both the harmonic components and influenza admission rates, the harmonic components were no longer predictive of HA-CDI rates. Conclusions: Harmonic regression provided a sensitive means of identifying seasonality in HA-CDI rates, but the seasonality effect was accounted for by influenza admission rates. The relationship between HA-CDI and influenza rates is likely mediated by antibiotic prescriptions, which needs to be assessed. To improve precision and reduce bias, research on interventions to reduce HA-CDI rates should assess historic seasonality in HA-CDI rates and should account for influenza admissions.Funding: NoneDisclosures: None