Forecasting Process

Step 2 Establish a forecast horizon, namely how far into the future we would like to make prediction. 

Step 4 Get past data and analyze. Validate the chosen forecasting technique. Step 5 Prepare the final forecast. 

Step 6 Monitor the forecasts regularly and make adjustments when needed. 

Step 3 is an important and difficult step in the forecasting process. There are a number of forecasting methods that are available in the literature. They can be broadly classified into two categories  

---

The practice of brand plan development and its NCE forecast component is designed to leverage knowledge from a cross-functional team, and in a certain sense it does. But because the traditional forecasting process is fundamentally disconnected from the operational realities of the disease marketplace, the marketer/brand manager lacks a tool to quantitatively analyze the effects of the product strategies developed. In other words. 

After the forecasting process the CHECK unit is istantiated to compare the estimated corrosion rate values with the measured one if a match of the values occours the current goal of the system is reached and the expert system waits until a new corrosion rate value is available. Otherwise the DIAGNOSIS units is istantiated. 

ANN, and apply it to the forecasting process of mine gas emission. The results show that, the model is feasible and effective, and can be used to predict the gas emission dangerous for large-scale coal seam By predicting that if the gas emission abnormal, we can take the following measures to control ... 

In the application of Grey System Theory to forecast the emission volume of gas from coal mine, we usually select a grey system theoretical differential equation model called GM(1, 1) which includes a first-order variable. The forecasting process can be split to three steps grey generation, parameter calculation and precision test. 

To provide a flexible, robust and effective methodology for fashion sales forecasting, this chapter examines the sales forecasting problem based on the forecasting process in real-world fashion retailing, which forecasts the total sales amount of each fashion item category or each city (all categories) on a medium-term basis (annually, quarterly and monthly). 

The change in demand forecasting processes is a major change management hnrdle for the traditional supply chain. The shift from a focus on history to a focus on market drivers, or to align on demand assumptions versus debating nnmbers, is a cnltural redefinition. 

Stage 1 The supply chain drives the efficient response. In this stage, the supply chain focuses on efficiency or asset utilization. The forecasting process is nascent. Companies model demand using basic forecasting technologies based on shipments and order data as inputs. The goal is to forecast future shipments. The data model represents what a manufacturer should make or a retailer should buy. 

In each stage of the market-driven capability model, the role of demand changes requiring a redefinition of forecasting processes. Additionally, at each stage of the demand process, a supply chain leader can chose to constrain the forecast (reducing volume requirements based on channel or supply-side constraints) to better manage the supply chain. 

Item segmentation by product forecastability helps companies to achieve greater accuracy in their product forecasting processes. An assessment helps to pinpoint the issues. The framework in Figure 3.5 is a useful way to segment products and apply the concepts to supply chain design. 

For many, this is a radical shift. Most demand forecasting processes are supply driven with little emphasis on predicting unconstrained demand, let alone shaping demand. In the interviews of 75 supply chain pioneers, when we asked about demand shaping, we found that most companies shift, versus shape, demand to meet supply constraints. Today, the norm is fitting demand to supply, rather than supply to demand. 

To be effective, the process needs discipline. The overall market-driven forecasting process design combines quantitative analysis... 

Market-driven demand management is supported by demand-driven forecasting principles that have a significant impact on a company, whether the company sells products or services. Companies that have implemented a market-driven forecasting process have experienced four key benefits ... 

Introduce S OP processes based on a strong demand-driven forecasting process that focuses on data and analytics to sense demand signals, and to shape and to translate demand to create a more accurate demand response. 

Visualize the goals of forecasting process improvement (the should-be state). 

Limited amount of data available for analysis or to be used in the forecast process (usually, sales data is available only for regular products)... 

No forecast tool is in place to automate statistical forecast process. 

Performance of planners and forecast are tied to compensation and rewards. Senior management understands, support and value demand planning function. Forecast tools are in place for both short and long term forecast, and are used to automate statistical forecast process, increasing planners capability to simulate different models. 

Demand signals are not considered in the forecast process, and forecasts are not communicated to other supply chain partners. 

Limited amount of data available for analysis or to be used in the forecast process. 

The forecasting process for end demand was continued on a routine basis with a monthly update and review. Any change in the level of demand was communicated to all parties in the system. As it turned out, the demand was stable for the next year but when it did drop off, at the end of the cycle, a new rate was set and all parties in the system adjusted to the new level. 

Mean Absolute Deviation (MAD) The average of absolute values of the deviations between observed and expected values. MAD can be calculated to evaluate forecasting processes such as the difference between actual sales and forecasts. 

The information structure In the basic form of the supply chain, day-to-day exchange of information is not possible, except for order placement. On the other hand, in a collaborative forecasting setting, sharing of forecast data and inventory status may be possible. Some of the models we shall present later are quite flexible in terms of their capability of handling a variety of information structures. We shall generally assume that a fair amount of static information is commonly known to the members of the supply chain. This includes the cost structure of the supply chain, the characteristics of the demand process, the types of forecasting processes used at each member s location. 

In this section we briefly describe a linear state space model that serves as a building block for the main models of collaborative forecasting processes we present in this chapter. We then present a well-known forecasting technique associated with this model namely, the Kalman filter. Let Xt be a finite, n-dimensional vector process called the state of the system. In the context of inventory management, this vector may consist of early indicators of future demand in the channel, actual demand realizations at various points of the channel, and so forth. Suppose that the state vector evolves according to ... 

---