Trend Models

There are three main types of trend models that are helpful to forecast data on past basis. These are linear trend model, exponential trend model and quadratic trend model. Linear trend model has two advantages that are more informative model in nature and has greater power in functional relationship (Anderson, 2001). Assignment Help Calculation of new data is a disadvantage of linear trend models because it has to be done repetitively. Exponential trend model is an important model that has simplicity to use and it is most economical model to collect data. The main disadvantage of the exponential trend model is its excessive simplicity and inflexibility (Brooks & Tsolacos, 2010). Finally, quadratic trend model has certain advantages over the equation because it has fulfillment based econometric criteria such as random distribution, autocorrelation etc. and the disadvantage of quadratic trend model is its instability in open ended category that may be restricted by these categories.


Criteria of Selection of Trend Model

There are four types of criteria that are used to select a trend model for the purpose of forecasting in an organization. Akaike Information Criteria (AIC) is a criterion that selects a nested econometric model. It is a method that solves the model selection problem by giving a mathematical framework (Leung, 2009). It prefers a model that has fewer parameters, so that it can control over increasing unnecessary parameters. Schwarz Information Criteria (SIC) is another criterion that selects a formal econometric model. It has higher penalty in comparison to AIC, so SIC is more appropriate to select the model. Hannan Quinn Information Criteria is use to handicap or penalize performance of a model for its different types of parameters. Option Criteria is a specify model selection criterion that is used to offer options given in the forecast statement (SAS Publishing, 2007).

Suitable Data for Linear Trend Model

A linear trend model is best fit for the simple linear data sets because linear trade line shows that some data values are increasing and decreasing in the graph with a stable rate. In this trend, several data can be used but simple data such as sales of a particular product for different time period is best fit in this model. On the other hand, linear trend model is best model for simple data as it shows relationship between independent and dependent variables that are visualized in the graph paper (Albright, Winston & Zappe, 2008). In linear trend model, researchers use data with their linear function that are estimated from the parameters, so these functions are called as parameters of the model. This model is also fit for simple data as it can use positive and negative both values.

Suitable Data for Exponential Trend Model

Exponential trend model is most suitable model for the data, which values are falling or rising at higher increasing rate. On the other hand, this model cannot use both positive and negative values because it can use only positive values but not zero or negative data values (Anderson, Sweeney & Williams, 2010). It is fit for positive data values because it can compute trend equation without using logarithm and regression analysis that is quite effective for the forecaster, while forecasting. At the same time, this trend model has capability to compute exponential trend equation that cannot be solved by using excel.


Suitable Data for Quadratic Trend Model

Quadratic trend model is used in upward trend data and their values do not appear in the trend equation. These values are estimated and assumed by regression model in the equation. It is best fit for these types of data because these types of trend equations might not be solved by linear trend models (Ragsdale, 2010). This model involves an equation that estimates a variable that is function of other variables like x is function of y but it requires two independent variables in place of one such as y1 and . It is also an appropriate method because upward trend can be followed by downward trend and vice - versa.

---

Blog Post Provided By - Homework Help Experts
Need Homework Help - Info@homeworkhelpexperts.com

• Call: UK Toll free no - 0044-808-189-1311t
• USA/CA Toll free no - 001-877-839-9989