Sunday, April 28, 2019
APPLIED DECISION METHODS CLASS; MOD4; Regression Models discussion Essay
APPLIED DECISION METHODS CLASS MOD4 Regression Models discussion - try on ExampleSales, and Price vs. Sales determined. Using computer software or excel, the coefficients can be generated and measures of uprightness of fit determined hence assisting in drawing the scatter diagrams.Render et al. asserts that assessing the plots (measured and calculated values of the hooked variable) and can be used to determine the quality of the model (pg. 127). For a good and apposite model, the trends should not be very dissimilar, and if the difference between measured and calculated points is huge, an ostensible trend should exist. Whats more, in the residual plot, the residual values of the dependent variable should be arbitrarily scattered around the logical argument err=0 and the mean should be zero (normality). For the model to be statistically valid and stable, the agency intervals to be used should be significantly small. An awkward model yields inaccurate derivative values and aerifo rm results for even smallest extrapolations. Additionally, for an appropriate model, the errors should have a constant variance that expedites the
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment