877. Trend Analysis – Why and How
$20.00
SAWE Members get a $200 store credit each year.*
*Store credit coupon available at checkout, click the button in your shopping cart to apply the coupon.
Not applicable to SAWE textbooks and current conference technical papers.
Paper
Abstract
This paper will discuss some of the basic philosophy behind and methods used in performing a trend analysis. Current problems associated with the management-weight control interface relative to trend analysis will also be discussed.
Weight control personnel have a two-fold problem not significantly different from that of other professionals. Their problems are the selling of services while keeping abreast of the current technical knowledge. The tendency by management and weight control personnel to ignore new methods and functions of mass property analysis must be overcome if weight control is to survive. No professional can afford the luxury of not increasing his knowledge and capability. Research is a never ending task which always must be done in conjunction with a close scrutiny of pertinent information. An emphasis should be placed upon quality.
In light of the current economic situation, weight control personnel must be more than clerks and calculators if their professional status is to be improved. One purpose of this paper is to give weight control personnel an approach which demonstrates to management the importance of an accurate and well-founded trend analysis.
As presented, the trend analysis has six major components as follows:
(1) Trending models
(2) Historical data
(3) Mass properties data base
(4) Projected weight growth pattern
(5) Potential changes
(6) Predicted outcome
All of these components have a role to play in the generation of a trend analysis. Special emphasis will be placed upon the mathematical aspect of trending models because this is normally the weakest link in the trend analysis process.
There are many mathematical approaches available. Trending models developed from mathematical tool such as linear and non-linear maximum likelihood equations, adaptive Fourier exponential equations, asymptotic exponential equations, and polynomial regression analysis have been are currently being used.
Due to time/space/presentation limitations, only the polynomial regression analysis and non-linear maximum likelihood equations will be developed relative to the mathematical trending curve procedures. A comparison of trending models and trend analyses at different time intervals in a design and production program will also be included.
