Forecasting

Abstract

Companies forecast in different ways and for different reasons. For the sake of my current employer, some of the reasons to why they forecast is to ensure that there are plenty of cellular phones in stock or even to make sure that the company has enough numbers to assign to customers when they purchase wireless service from us. The following paper will compare and contrast various forecasting methods while also elaborating on the method that my current employer use for forecasting sales and mobile identification number (MDN) requirements.

Forecasting Assignment

Forecasting is the ability to plan ahead for future expectations of what the future may hold. For example, business forecast every year for what they feel that particular company should accomplish. A sales department forecasts how many sales not only the department should do as a hole but how many sales each individual sales representative should also do. Every year, meteorologists forecast the number of hurricanes that may or may not hit the United States as well as other countries. The above leaves me to ask, what is forecasting? In a lack of words, forecasting is predicting what may or may not happen in the future. The forecast could simply be about sales of automobiles, wireless services or the number and strengths of hurricanes. Forecasting is defined as "to calculate or predict some future event or condition, as a result of study and analysis of pertinent data." (15c, Unknown) There are four basic types of forecasting qualitative, casual relationships, simulation and time series analysis. The following paper will compare and contrast forecasting methods and discuss the current forecasting methods of my current employer.

As previously stated, I will compare and contrast several forecasting methods. Prior to doing so, I will first define each forecasting method in order to give the reader a better understanding of each. The first method that I will define is Delphi. The Delphi method for forecasting can be found under the Qualitative type of forecasting since the utilization of this method focuses on judgments and are based on estimate and opinions."The Delphi method is a systematic interactive forecasting method based on independent inputs of selected experts." (Wales, 2006) This method's name comes from the Oracle of Delphi and was introduced during 1950 to 1960's by Olaf Hemer and Norman Dalkey. This approach is formed by gathering various questions that are formed as hypotheses and posed to a group of experts to answer each and state when he or she believes the hypotheses will be fulfilled.

"A time series is a sequence of observations of a periodic random variable. Time series are important for operations research because they are often the drivers of decision models." (Jensen, 2004) Belief is that the Time series analysis can construct from previous history in order to predict the future. This forecasting method is primarily used for making decisions in the future such as ordering supplies or employment. Additional examples of tasks that may require this type of forecasting would be an inventory control system, course scheduling or a university for staffing for future student inflow, just to name a few. The Shiskin time series forecast is one of many under Time series analysis. Julius Shiskin of the Census Bureau developed this forecast to be used to track the turning points for a company. For this to be effective, one should have three years of history available.

To recap, I have discussed the Delphi and Time series forecasting methods. Next I will discuss Input/output forecasting method. Input/Output forecasting is a method that can be found under Trend projections. In Trend projections one "fits a mathematical trend line to the data points and projects it into the future."(Chase, etc. 2005) The Input/output forecasting technique compares the sales of companies within an industry to other companies and the government as well. The changes of what an industry that produces may see in the future due to the buying changes of another.

The final type of forecasting to be discussed is Casual relationship forecasting. Casual relationship is "a situation in which one event causes another. If the event is far enough in the future, it can be used as a basis for forecasting." A good example of this would be a casual relationship between student enrollment and teachers in a university. If the enrollment of students is higher than normal than it would cause for additional teachers to be hired and vice versa. This is considered a casual relationship since additional teachers would not have been hired if enrollment was not up therefore the high enrollment causes the increased staff.

Now that I have defined each of the forecasting methods discussed within this paper, I will now compare them to each other. Of the methods discussed in this paper, Delphi has no comparison to any of the other three methods of forecasting. Delphi forecasting is used in conjunction with experts and hypotheses rather than gathering information from other sources and analyzing them. At the same time, Time series and Casual relationship forecasting has similarities because each is dependent on another aspect. Though Time series focuses on the seasonal aspects of things, casual relationship could very well be dependent

on the seasons. For example, the Time series forecasting is big on farmers for planning their crops; however bad weather would cause a casual relationship between the two. This reasoning is due to the fact that the forecast previously given for crops via Time series is dependent