Linear model assumptions - ANSWER Linearity, independence, normality,
homoscedasticity
Problems with correlated errors in linear model - ANSWER Standard errors
underestimate true errors, CIs and PIs narrower, p-values lower
Effects of heteroscedasticity - ANSWER Estimates no longer BLUE, standard
errors, CIs, and hypothesis tests invalid
Outlier - ANSWER A response value that is far from the value predicted by the
model
Leverage point - ANSWER A value of a predictor variable that is far from the
rest of the x values
Effects of collinearity - ANSWER Hard to interpret individual coefficients,
larger standard errors, lower t-statistics, reduced power
Why use variance inflation factor (VIF)? - ANSWER Better detects
multicollinearity, i.e., when relationships are not limited to pairwise
VIF definition - ANSWER Ratio of the variance of beta_j when fitting the full
model divided by the variance of beta_j if fit on its own
Which data points do forecasts weight the most? - ANSWER The most recent
ones
Weakly stationary - ANSWER Constant mean and constant variance; the mean
of y does not depend on t, and Cov(y_t, y_s) depends only on | t - s |
Why is white noise both the least and most important model? - ANSWER Least
important because it assumes no relationship between points; most important
because most models try to reduce a time series to white noise
What type of model is a differenced random walk model? - ANSWER White
noise
When to use differences of logs in random walks - ANSWER When both the
series variance and log series variance increase over time
Interpretation of lag k partial autocorrelation - ANSWER Correlation between
, y_t and y_{t-k}, controlling for the effects of the intervening variables (y_{t-1}, ... ,
y_{t-k+1})
Odds ratio interpretation for binary X - ANSWER The odds when x = 1 are
exp(beta_j) greater than the odds when x = 0
Odds ratio interpretation for continuous X - ANSWER beta_j is the the
proportional change in the odds ratio
In GLMs, what drives many inference properties? - ANSWER The choice of
variance function, not the choice of the distribution
Usefulness of R squared in nonlinear models - ANSWER R squared is not a
useful statistic in nonlinear models, in part because the ANOVA decomposition is
no longer valid
Why use Anscombe residuals? - ANSWER Transform to normal distribution or
to stabilize variance
Properties of deviance residuals - ANSWER Similar to Anscombe for approx.
normality, readily defined for any GLM model, easy to compute
In classification trees, what measure(s) is (are) used for tree growing and why? -
ANSWER Gini index and entropy; they are more sensitive to node purity
When is classification error rate preferable in classification trees? - ANSWER
When pruning the tree and prediction accuracy of the final pruned tree is most
important
Advantages of decision trees - ANSWER Easy to explain (easier than linear
regression), mimic human decision-making, graphic displays, handle qualitative
predictors without the needs to create dummy variables
Disadvantages of decision trees - ANSWER Less predictive accuracy,
non-robust
Why use OOB estimation? - ANSWER Directly estimates test error and is
convenient when performing bagging on large data sets for which CV is
computationally difficult
Advantages of bagging over regular decision tree - ANSWER Variance
reduction and better accuracy
Parameters in boosting trees - ANSWER Number of trees, shrinkage
Voordelen van het kopen van samenvattingen bij Stuvia op een rij:
Verzekerd van kwaliteit door reviews
Stuvia-klanten hebben meer dan 700.000 samenvattingen beoordeeld. Zo weet je zeker dat je de beste documenten koopt!
Snel en makkelijk kopen
Je betaalt supersnel en eenmalig met iDeal, creditcard of Stuvia-tegoed voor de samenvatting. Zonder lidmaatschap.
Focus op de essentie
Samenvattingen worden geschreven voor en door anderen. Daarom zijn de samenvattingen altijd betrouwbaar en actueel. Zo kom je snel tot de kern!
Veelgestelde vragen
Wat krijg ik als ik dit document koop?
Je krijgt een PDF, die direct beschikbaar is na je aankoop. Het gekochte document is altijd, overal en oneindig toegankelijk via je profiel.
Tevredenheidsgarantie: hoe werkt dat?
Onze tevredenheidsgarantie zorgt ervoor dat je altijd een studiedocument vindt dat goed bij je past. Je vult een formulier in en onze klantenservice regelt de rest.
Van wie koop ik deze samenvatting?
Stuvia is een marktplaats, je koop dit document dus niet van ons, maar van verkoper luzlinkuz. Stuvia faciliteert de betaling aan de verkoper.
Zit ik meteen vast aan een abonnement?
Nee, je koopt alleen deze samenvatting voor €9,30. Je zit daarna nergens aan vast.