Handbook of Regression Analysis With Applications in R. Samprit Chatterjee

Handbook of Regression Analysis With Applications in R

Handbook of Regression Analysis With Applications in R - Samprit Chatterjee

Скачать книгу
accounted for by the best linear combination of the predictors. Values closer to
indicate a good deal of predictive power of the predictors for the target variable, while values closer to
indicate little predictive power. An equivalent representation of
is

      where

      is the sample correlation coefficient between

and
(this correlation is called the multiple correlation coefficient). That is,
is a direct measure of how similar the observed and fitted target values are.

      It can be shown that

is biased upwards as an estimate of the population proportion of variability accounted for by the regression. The adjusted
corrects this bias, and equals

is large relative to
(that is, unless the number of predictors is large relative to the sample size),
and
will be close to each other, and the choice of which to use is a minor concern. What is perhaps more interesting is the nature of
as providing an explicit tradeoff between the strength of the fit (the first term, with larger
corresponding to stronger fit and larger
) and the complexity of the model (the second term, with larger
corresponding to more complexity and smaller
). This tradeoff of fidelity to the data versus simplicity will be important in the discussion of model selection in Section 2.3.1.

      The only parameter left unaccounted for in the estimation scheme is the variance of the errors

. An unbiased estimate is provided by the residual mean square,

really say anything of value about
? This isn't a question that can be answered completely statistically; it requires knowledge and understanding of the data and the underlying random process (that is, it requires context). Recall that the model assumes that the errors are normally distributed with standard deviation
. This means that, roughly speaking,
of the time an observed
value falls within
of the expected response

      

can be estimated for any given set of
values using

that can be used in constructing this rough prediction interval
.

      1.3.3 HYPOTHESIS TESTS AND CONFIDENCE INTERVALS FOR β

      There are two types of hypothesis tests of immediate interest related to the regression coefficients.

      1 Do any of the predictors provide predictive power for the target variable? This is a test of the overall significance of the regression,versusThe test of these hypotheses is the ‐test,This is referenced against a null ‐distribution on degrees of freedom.

      2 Given the other variables in the model, does a particular predictor provide additional predictive power? This corresponds to a test of the significance of an individual coefficient,versusThis is tested using a ‐test,which is compared to a ‐distribution on degrees of freedom. Other values of can be specified in the null hypothesis (say ), with the ‐statistic becoming(1.9) The values of are obtained as the square roots of the diagonal elements of , where is the residual mean square (1.8). Note that for simple regression (), the hypotheses corresponding to the overall significance of the model and the significance of the predictor are identical,versusGiven the equivalence of the sets of hypotheses, it is not surprising that the associated tests are also equivalent; in fact, , and the associated tail probabilities of the two tests are identical.A ‐test for the intercept also can be constructed as in (1.9), although this does not refer to a hypothesis about a predictor, but rather about whether the expected target is equal to a specified value if all of the predictors equal zero. As was noted in Section 1.3.1, this is often not physically meaningful (and therefore of little interest), because the condition that all predictors equal zero cannot occur, or does not come close to occurring in the observed data.

      As

Скачать книгу