introducción al análisis de la regresión multivariante

Regression analysis is a method for exploring the relationship between a response variable and one or more explanatory variables..

The basic multiple regression model assumes a linear relationship between x_i and y, with superimposed noise (error) e

Y = b₀ + b₁x₁ + b₂x₂ + b₃x₃ + e     (1)

Using certain statistical assumptions about (1), we can "fit" a regression equation

y = b₀ + b₁x₁ + b₂x₂ + b₃x₃              (2)

as an estimator of the true relationship between the x_is and y

y = b₀ + b₁x₁ + b₂x₂ + b₃x₃          (3)

We won't claim that multiple regression is an easy subject; however, the Multiple Linear Regression components from Applied Analytic Systems allow Delphi programmers to add MLR capabilities to their applications with ease. You don't need a degree in statistics, nor do you need to understand the underlying statistical mathematics in order to make regression a valuable part of your toolset.

This tutorial will demonstrate the use of TMLRegress by developing, start-to-finish, a simple Delphi 2.0 application to analyze the data collected concerning the age and maintenance record of vehicles. After constructing and solving the model, we will comment on the validity of the model and some of the implications of the analysis.

TMLRegress Analytic Systems Inc. Delphi, Borland International, Inc. 19.may.2000 -