Search results

Demand forecasting has long been an imperative tenet in production planning especially in a make-to-order environment where a typical manufacturer has to balance the issues of holding excessive safety stocks and experiencing possible stockout. Many studies provide pragmatic paradigms to generate demand forecasts (mainly based on smoothing forecasting models.) At the same time, artificial neural networks (ANNs) have been emerging as alternatives. In this chapter, we propose a two-stage forecasting approach, which combines the strengths of a neural network with a more conventional exponential smoothing model. In the first stage of this approach, a smoothing model estimates the series of demand forecasts. In the second stage, general regression neural network (GRNN) is applied to learn and then correct the errors of estimates. Our empirical study evaluates the use of different static and dynamic smoothing models and calibrates their synergies with GRNN. Various statistical tests are performed to compare the performances of the two-stage models (with error correction by neural network) and those of the original single-stage models (without error-correction by neural network). Comparisons with the single-stage GRNN are also included. Statistical results show that neural network correction leads to improvements to the forecasts made by all examined smoothing models and can outperform the single-stage GRNN in most cases. Relative performances at different levels of demand lumpiness are also examined.

Research in the area of forecasting and stock inventory control for intermittent demand is designed to provide robust models for the underlying demand which appears at random, with some time periods having no demand at all. Croston’s method is a popular technique for these models and it uses two single exponential smoothing (SES) models which involve smoothing constants. A key issue is the choice of the values due to the sensitivity of the forecasts to changes in demand. Suggested selections of the smoothing constants include values between 0.1 and 0.3. Since an ARIMA model has been illustrated to be equivalent to SES, an optimal smoothing constant can be selected from the ARIMA model for SES. This chapter will conduct simulations to investigate whether using an optimal smoothing constant versus the suggested smoothing constant is important. Since SES is designed to be an adapted method, data are simulated which vary between slow and fast demand.

We investigate whether smooth robust methods for forecasting can help mitigate pronounced and persistent failure across multiple forecast horizons. We demonstrate that naive predictors are interpretable as local estimators of the long-run relationship with the advantage of adapting quickly after a break, but at a cost of additional forecast error variance. Smoothing over naive estimates helps retain these advantages while reducing the costs, especially for longer forecast horizons. We derive the performance of these predictors after a location shift, and confirm the results using simulations. We apply smooth methods to forecasts of UK productivity and US 10-year Treasury yields and show that they can dramatically reduce persistent forecast failure exhibited by forecasts from macroeconomic models and professional forecasters.

The work of Scott, Bruce and Cooper on small firm growth and development is reviewed. It is shown that by adapting exponential smoothing forecasting procedures it is possible to monitor the commercial health of a small firm. This is achieved by ‘tracking’ key indicators and producing an exception message when a signal exceeds certain predetermined control limits. The procedure is equally effective for either a step or ramp change in the underlying input data. This suggested approach requires little sophistication in either data or technique and has a practical application to small firm management, while adding to our understanding of the process of growth of small businesses.

To provide an overview of how a number of frequently used smoothing‐based forecasting techniques can be modelled for use in dynamic analysis of production‐inventory systems.

The smoothing techniques are modelled using transfer functions and state space representation. Basic control theory is used for analysing the dynamic properties.

A set of expressions are derived for the smoothing techniques and dynamic properties are identified.

Dynamic properties are important in many applications. It is shown that the different smoothing techniques can have very different influences on the dynamic behaviour and therefore should be considered as a factor when smoothing parameters are decided on.

Dynamic behaviour of production‐inventory systems can be analysed using control theory based on, e.g. transfer functions or state space models. In this paper a set of models for five common smoothing techniques are analysed and their respective dynamic properties are highlighted.

When forecasting intermittent demand the method derived by Croston (1972) is often cited. Previous research favorably compared Croston's forecasting method for demand with simple exponential smoothing assuming a nonzero demand occurs as a Bernoulli process with a constant probability. In practice, however, the assumption of a constant probability for the occurrence of nonzero demand is often violated. This research investigates Croston's method under violation of the assumption of a constant probability of nonzero demand. In a simulation study, forecasts derived using single exponential smoothing (SES) are compared to forecasts using a modification of Croston's method utilizing double exponential smoothing to forecast the time between nonzero demands assuming a normal distribution for demand size with different standard deviation levels. This methodology may be applicable to forecasting intermittent demand at the beginning or end of a product's life cycle.

Weak-form rationality of fixed-event forecasts implies that forecast revisions should not be correlated. However, significant positive correlations between consecutive forecast revisions were found in most USDA forecasts for U.S. corn, soybeans, wheat, and cotton. This study developed a statistical procedure for correction of this inefficiency which takes into account the issue of outliers, the impact of forecast size and direction, and the stability of revision inefficiency. Findings suggest that the adjustment procedure has the highest potential for improving accuracy in corn, wheat, and cotton production forecasts.

This chapter uses advance order data from an actual manufacturing shop to develop and test a forecast model for total demand. The proposed model made direct use of historical time series data for total demand and time series data for advance orders. Comparison of the proposed model to commonly used approaches showed that the proposed model exhibited greater forecast accuracy.

This study examines the use of forecast combination to improve the accuracy of forecasts of cumulative demand. A forecast combination methodology based on least absolute value (LAV) regression analysis is developed and is applied to partially accumulated demand data from an actual manufacturing operation. The accuracy of the proposed model is compared with the accuracy of common alternative approaches that use partial demand data. Results indicate that the proposed methodology outperforms the alternative approaches.

The purpose of this paper is to propose a forecasting model to predict the demand under uncertain environment to control the bullwhip effect (BWE) considering review-period order-up-to level ((R, S)) inventory control policy and its different variants such as (R, βS) (R, γO) and (R, γO, βS) proposed by Jakšič and Rusjan, (2008) and Bandyopadhyay and Bhattacharya (2013).

A hybrid forecasting model has been developed by combining the feature of discrete wavelet transformation (DWT) and an intelligence technique, multi-gene genetic programming (MGGP), denoted as DWT-MGGP. Performance of DWT-MGGP model has been verified under (R, S) inventory control policy considering demand from three different manufacturing companies.

A comparison between DWT-MGGP model and autoregressive integrated moving average forecasting model has been done by estimating forecast error and BWE. Further, this study has been extended with analysing the behaviour of BWE considering different variants of (R, S) policy such as (R,βS) (R, γO) and (R,γO,βS) and found that BWE can be moderated by controlling the inventory smoothing (β) and order smoothing parameters (γ).

This study is limited to different variants of (R, S) inventory control policy. However, this study can be further extended to continuous review policy.

The proposed DWT-MGGP model can be used as a suitable demand forecasting model to control the BWE when (R, S), (R,βS) (R,γO) and (R,γO,βS)inventory control policies are followed for replenishment.

This study analyses the behavior of BWE through controlling the inventory smoothing (β) and order smoothing parameters (γ) when demand is predicted using DWT-MGGP forecasting model and order is estimated using (R, S), (R,βS) (R,γO) and (R,γO,βS) inventory control policies.