Free Practice Questions for SAS Statistics for Machine Learning

Statistics and Machine Learning

Foundational concepts that distinguish statistical inference from machine-learning prediction, supervised vs. unsupervised learning, and sampling and partitioning strategies.

Fundamental Statistical Concepts

Descriptive statistics with PROC MEANS / UNIVARIATE / FREQ, probability distributions, hypothesis tests in PROC TTEST and PROC NPAR1WAY, one-way and factorial ANOVA via PROC GLM with Type I/II/III SS and Tukey HSD, plus Pearson and Spearman correlation in PROC CORR.

Explanatory Modeling Using Linear Regression

PROC REG and PROC GLM, residual and influence diagnostics (Cook's D, leverage, DFBETAS, studentized residuals), multicollinearity (VIF, condition index), categorical predictors and interactions, and stepwise / LASSO / elastic-net selection in PROC GLMSELECT.

Predictive Modeling Using Logistic Regression

PROC LOGISTIC for binary, ordinal proportional-odds, and nominal multinomial models; odds ratios, ROC curves, c-statistic, Hosmer-Lemeshow, deviance residuals, FIRTH for separation, and rare-event handling.

Statistical Foundations of Machine Learning

Bias-variance tradeoff, K-fold and stratified cross-validation, regularization (LASSO, ridge, elastic net) via PROC GLMSELECT and PROC HPGENSELECT, decision trees in PROC HPSPLIT, PCA via PROC PRINCOMP, and clustering with PROC FASTCLUS.