Nonlinear Regresion Models
Michael S Lee
The following labs are from James, Witten, Hastie, and Tibshirani’s An Introduction to Statistical Learning, pp. 115-159.
We load the libraries to be used.
library(ISLR); library(MASS); data(Boston)Non-linear Transformation of the Predictors
Non-linear transformation of individual predictor variables is largely akin to the interaction of different predictor variables to create another predictor, so the placement of this portion of the tutorial is somewhat arbitrary.
lm.fit2 <- lm(medv~lstat+I(lstat^2),data=Boston)
summary(lm.fit2)##
## Call:
## lm(formula = medv ~ lstat + I(lstat^2), data = Boston)
##
## Residuals:
## Min 1Q Median 3Q Max
## -15.2834 -3.8313 -0.5295 2.3095 25.4148
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 42.862007 0.872084 49.15 <2e-16 ***
## lstat -2.332821 0.123803 -18.84 <2e-16 ***
## I(lstat^2) 0.043547 0.003745 11.63 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 5.524 on 503 degrees of freedom
## Multiple R-squared: 0.6407, Adjusted R-squared: 0.6393
## F-statistic: 448.5 on 2 and 503 DF, p-value: < 2.2e-16We analyze the variance of the dataset via the anova() function, i.e., we use anova() to quantify the extent to which the quadratic fit outperforms the linear fit.
lm.fit <- lm(medv~lstat,data=Boston)
anova(lm.fit,lm.fit2)## Analysis of Variance Table
##
## Model 1: medv ~ lstat
## Model 2: medv ~ lstat + I(lstat^2)
## Res.Df RSS Df Sum of Sq F Pr(>F)
## 1 504 19472
## 2 503 15347 1 4125.1 135.2 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1We produce some plots to evaluate the lm.fit2.
par(mfrow=c(2,2))
plot(lm.fit2)
We also examine the fit quality of higher-order transformations.
lm.fit5<-lm(medv~poly(lstat,5), data=Boston)
summary(lm.fit5)##
## Call:
## lm(formula = medv ~ poly(lstat, 5), data = Boston)
##
## Residuals:
## Min 1Q Median 3Q Max
## -13.5433 -3.1039 -0.7052 2.0844 27.1153
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 22.5328 0.2318 97.197 < 2e-16 ***
## poly(lstat, 5)1 -152.4595 5.2148 -29.236 < 2e-16 ***
## poly(lstat, 5)2 64.2272 5.2148 12.316 < 2e-16 ***
## poly(lstat, 5)3 -27.0511 5.2148 -5.187 3.10e-07 ***
## poly(lstat, 5)4 25.4517 5.2148 4.881 1.42e-06 ***
## poly(lstat, 5)5 -19.2524 5.2148 -3.692 0.000247 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 5.215 on 500 degrees of freedom
## Multiple R-squared: 0.6817, Adjusted R-squared: 0.6785
## F-statistic: 214.2 on 5 and 500 DF, p-value: < 2.2e-16The summary suggests that including additional, higher-order polynomials, at least up to the fifth order, improves the fit.
We can also transform predictors in other ways, i.e., we’re not limited to simple higher-order versions of predictors, but can apply, for instance, a log transformation.
summary(lm(medv~log(rm),data=Boston))##
## Call:
## lm(formula = medv ~ log(rm), data = Boston)
##
## Residuals:
## Min 1Q Median 3Q Max
## -19.487 -2.875 -0.104 2.837 39.816
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -76.488 5.028 -15.21 <2e-16 ***
## log(rm) 54.055 2.739 19.73 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 6.915 on 504 degrees of freedom
## Multiple R-squared: 0.4358, Adjusted R-squared: 0.4347
## F-statistic: 389.3 on 1 and 504 DF, p-value: < 2.2e-16Qualitative Predictors
We now leave the Boston dataset and examine the Carseats data. The objective is to predict “Sales” in 400 places based on a number of predictors.
data(Carseats)
names(Carseats)## [1] "Sales" "CompPrice" "Income" "Advertising" "Population"
## [6] "Price" "ShelveLoc" "Age" "Education" "Urban"
## [11] "US"The Carseats dataset includes qualitative, or class, variables, e.g., the “Shelveloc,” which indicate shelving quality for the merchandise. The following model fits a multiple regression, which includes interaction terms.
lm.fit=lm(Sales~.+Income:Advertising+Price:Age,data=Carseats)
summary(lm.fit)##
## Call:
## lm(formula = Sales ~ . + Income:Advertising + Price:Age, data = Carseats)
##
## Residuals:
## Min 1Q Median 3Q Max
## -2.9208 -0.7503 0.0177 0.6754 3.3413
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 6.5755654 1.0087470 6.519 2.22e-10 ***
## CompPrice 0.0929371 0.0041183 22.567 < 2e-16 ***
## Income 0.0108940 0.0026044 4.183 3.57e-05 ***
## Advertising 0.0702462 0.0226091 3.107 0.002030 **
## Population 0.0001592 0.0003679 0.433 0.665330
## Price -0.1008064 0.0074399 -13.549 < 2e-16 ***
## ShelveLocGood 4.8486762 0.1528378 31.724 < 2e-16 ***
## ShelveLocMedium 1.9532620 0.1257682 15.531 < 2e-16 ***
## Age -0.0579466 0.0159506 -3.633 0.000318 ***
## Education -0.0208525 0.0196131 -1.063 0.288361
## UrbanYes 0.1401597 0.1124019 1.247 0.213171
## USYes -0.1575571 0.1489234 -1.058 0.290729
## Income:Advertising 0.0007510 0.0002784 2.698 0.007290 **
## Price:Age 0.0001068 0.0001333 0.801 0.423812
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.011 on 386 degrees of freedom
## Multiple R-squared: 0.8761, Adjusted R-squared: 0.8719
## F-statistic: 210 on 13 and 386 DF, p-value: < 2.2e-16Use the contrasts() function with a categorical variable as the argument to return the coding that R uses in assigning dummy variables.
attach(Carseats)
contrasts(ShelveLoc)## Good Medium
## Bad 0 0
## Good 1 0
## Medium 0 1Notice the higher coefficient associated with ShelveLocGood, which R created, is higher than that of ShelveLocMedium, which makes sense since better shelving should lead to higher Sales.