Loading demo links...
What to study for each A00-485 domain
SAS Visual Statistics cross-functional tasks
This section covers the shared setup and workflow tasks used across all models in SAS Visual Statistics on SAS Viya. It includes loading data, assigning data roles, applying filters, creating partitions, and using common exploration and model options.
Key skills
data roles, response and predictor assignment, data source and object filters, data partitions, calculated items, binning, missing value handling, variable importance, model workflow
Building and assessing segmentation models
This section covers building segmentation models that group observations into clusters with similar profiles. It also covers assessing those segments using cluster sizes, centroids, and cluster profiles to judge separation and usefulness.
Key skills
clustering and segmentation concepts, k-means assignment, centroids, standardization, number of clusters, cluster ID, cluster summary and profiles, segment assessment
Modelling using decision trees
This section covers building decision trees for classification and prediction, including how splits are chosen and how growth is controlled. It also covers pruning, subtree assessment, and interpreting leaf predictions and variable importance.
Key skills
split criteria, entropy and Gini, gain and variance reduction, maximum depth, minimum leaf size, pruning, subtree assessment, misclassification, variable importance, leaf predictions
Building and assessing regression-type models
This section covers building and assessing regression-type models, including linear regression, logistic regression, and generalized linear models. It also covers variable selection, model effects, diagnostics, and assessment statistics for interval and binary targets.
Key skills
least squares, coefficients and intercepts, R-square and adjusted R-square, AIC and BIC, ASE and RMSE, logistic regression and odds ratios, GLMs and link functions, stepwise and LASSO selection, interactions, residuals, VIF, ROC and lift, confusion matrix, KS
Model comparison and scoring
This section covers comparing candidate models on common validation data using appropriate fit and prediction statistics to select a champion. It also covers scoring new data with a fitted model and checking scored output before use.
Key skills
model comparison, champion selection, validation statistics, scoring new data, predictor requirements for scoring, predicted values and event probabilities, scored output review