Advances in Business and Management Forecasting: Volume 9

The application of forecasting to health care is not new. A frequent issue in many Inpatient Rehabilitation Facilities (IRFs) is the fluctuating and unpredictable census. With scarce resources, particularly physical therapists and occupational therapists, this unpredictability makes appropriate scheduling of these resources challenging. This research addresses the issue of patient admissions in an inpatient rehabilitation facility attached to an 861 bed level-one trauma hospital. The goal is to develop a predictive model for the IRF's Census to assist in resource planning (e.g., labor, beds, and materials).

Evaluating pain and discomfort in animals is difficult at best. Veterinarians believe however, that they can establish a proxy for estimating levels of pain and discomfort in canines by observing variations in their activity levels. Sufficient research has been conducted to justify this assertion, but little has been conducted to analyze the volumes of activity data collected. We present the first of a series of analyses aimed at ultimately presenting an effective predictive tool for canine pain and discomfort levels. In this chapter, we perform analyses on a dataset of normal (control) dogs, containing almost 3 million records. The forecasting analyses incorporated multiple polynomial regression models with transcendental transformations and ARIMA models to provide effective determination and prediction of baseline normal canine activity levels.

Managing a large hospital network can be an extremely challenging task. Management must rely on numerous pieces of information when making business decisions. This chapter focuses on the number of bed days (NBD) which can be extremely valuable for operational managers to forecast for logistical planning purposes. In addition, the finance staff often requires an expected NBD as input for estimating future expenses. Some hospital reimbursement contracts are on a per diem schedule, and expected NBD is useful in forecasting future revenue.Two models, time regression and autoregressive integrated moving average (ARIMA), are applied to nine years of monthly counts of the NBD for the Rhode Island Hospital System. These two models are compared to see which gives the best fit for the forecasted NBD. Also, the question of summarizing the time data from monthly to quarterly time periods is addressed. The approaches presented in this chapter can be applied to a variety of time series data for business forecasting.

The service sector comprises a dominant segment of the economy. Customer satisfaction, a measure of quality, is based on the degree of difference between expected quality and the actual level of quality experienced. Expected level of quality is influenced by customer perception of quality, which in turn is impacted by external and internal factors. In service industries, the interaction between the service provider and the customer may also influence quality. Thus quality may consist of tangible and intangible factors. In this chapter we consider the measurable attributes associated with quality in the service sector. Based on a specified guarantee level associated with the attribute, for example, service time, a penalty function is used to determine the impact of deviating from the guarantee level. With service time being a stochastic random variable, expected penalty costs to the service provider are found under a variety of conditions.

In this chapter, we argue that under- and over-reaction are both parts of the price dynamics caused by investor's naïve judgmental extrapolation. We propose to use the Holt–Winters model, a parsimonious model with two parameters, to represent investor's conservatism (anchoring) and representativeness (trending). The complexity of earning information, which is broken down into a drift, a transitory shock, and an autocorrelated permanent shock, add further volatility to the price. We explain the price dynamics caused by the interplay of the earning model and investor's naïve belief. It is further argued that empirical “underreaction” and “overreaction” differ from true under- and overreaction. The simulated results with the proposed model confirm with empirical findings on under- and overreaction.

In this chapter, we demonstrate that studying relevant investment information helps reduce individual investors’ disposition effect. It is prevalent that many individual investors in stock market do not form their own opinion about the investments; instead they mimic investment strategies of others. This research shows that the intention of making easy money only worsens the disposition effect. We collect 2,632 individual stock investors through nationwide surveys in Taiwan. Using regression models, we examine the effects of study on reducing investors’ inclination of holding-losers/selling-winners and the disposition effect. The findings show that investors realize losses sooner and significantly reduce the disposition effect if they choose to learn about their investments. The results also demonstrate that if the investors are willing to learn about firms in which they invest, they become more rational about their investment decisions. They are no longer influenced by the sentiment of regret resistance or misperception of the stock trend, which in turn reduces the disposition effect. This study supports that investors make better investment decisions if they perform necessary due diligence prior to investing.

The purpose of this study is to analyze structural changes that took place in the cotton industry and develop a statistical model that reflects the current drivers of U.S. upland cotton prices. This study concludes that a structural break in the U.S. cotton industry occurred in 1999, and that world cotton supply has become an important determinant of U.S. cotton prices. The model developed here forecasts changes in U.S. cotton price based on changes in U.S. cotton supply, changes in U.S. stocks-to-use ratio (S/U), changes in China's net imports as a share of world consumption, the proportion of U.S. cotton engaged in the loan program, and changes in world supply of cotton.

This research examines the use of econometric models to predict the total net asset value (NAV) of an asset allocation mutual fund. In particular, the mutual fund case used is the Vanguard Wellington Fund (VWELX). This fund maintains a balance between relatively conservative stocks and bonds. The period of the study on which the prediction of the total NAV is based is the 24-month period of 2010 and 2011 and the forecasting period is the first three months of 2012. Forecasting the total NAV of a massive conservative allocation fund, composed of an extremely large number of investments, requires a method that produces accurate results. Achieving this accuracy has no necessary relationship to the complexity of the methods typically employed in many financial forecasting studies.

This chapter illustrates the Technology Forecasting using Data Envelopment Analysis (TFDEA) process on Liquid Crystal Display (LCD) performance characteristics from 1997 to 2012. The objective of this study is to forecast future state-of-the-arts (SOAs) specifications as well as to diagnose past technological advancement of the LCD industry. Appropriate characteristics were determined from a group of LCD technologists. Data was gathered from public databases and outlying data points were cross-referenced as a validity check. The TFDEA process is defined and its application to the dataset is described in detail. The results not only provide information on how LCD industry has evolved but also provide an insight on future NPD targets.

Retail sales usually exhibit strong trend and seasonal patterns. Practitioners have typically used seasonal autoregressive integrated moving average (ARIMA) models to predict retail sales exhibiting these patterns. Due to economic instability, recent retail sales time-series data show a higher degree of variability and nonlinearity, which makes the ARIMA model less accurate. This chapter demonstrates the feasibility and potential of applying empirical mode decomposition (EMD) in forecasting aggregate retail sales. The hybrid forecasting method of integrating EMD and neural network (EMD-NN) models was applied to two real data sets from two different time periods. The one-period ahead forecasts for both time periods show that EMD-NN outperforms the classical NN model and seasonal ARIMA. In addition, the findings also indicate that EMD-NN can significantly improve forecasting performance during the periods in which macroeconomic conditions are more volatile.

Past research has shown that forecast combination typically improves demand forecast accuracy even when only two component forecasts are used; however, systematic bias in the component forecasts can reduce the effectiveness of combination. This study proposes a methodology for combining demand forecasts that are biased. Data from an actual manufacturing shop are used to develop the methodology and compare its accuracy with the accuracy of the standard approach of correcting for bias prior to combination. Results indicate that the proposed methodology outperforms the standard approach.

One aspect of forecasting intermittent demand for slow-moving inventory that has not been investigated to any depth in the literature is seasonality. This is due in part to the reliability of computed seasonal indexes when many of the periods have zero demand. This chapter proposes an innovative approach which adapts Croston's (1970) method to data with a multiplicative seasonal component. Adaptations of Croston's (1970) method are popular in the literature. This method is one of the most popular techniques to forecast items with intermittent demand. A simulation is conducted to examine the effectiveness of the proposed technique extending Croston's (1970) method to incorporate seasonality.

This study examines the scheduling problem for a two-stage flowshop. All jobs are immediately available for processing and job characteristics including the processing times and due dates are known and certain. The goals of the scheduling problem are (1) to minimize the total flowtime for all jobs, (2) to minimize the total number of tardy jobs, and (3) to minimize both the total flowtime and the total number of tardy jobs simultaneously. Lower bound performances with respect to the total flowtime and the total number of tardy jobs are presented. Subsequently, this study identifies the special structure of schedules with minimum flowtime and minimum number of tardy jobs and develops three sets of heuristics which generate a Pareto set of bicriteria schedules. For each heuristic procedure, there are four options available for schedule generation. In addition, we provide enhancements to a variety of lower bounds with respect to flowtime and number of tardy jobs in a flowshop environment. Proofs and discussions to lower bound results are also included.