100+ Free SAS NLP and Computer Vision (Viya 3.5) Practice Questions

Loading and Exploring Data

Import documents into VTA, set project language, load images with image.loadImages, resize and rescale images, run augmentImages, and explore parsed term roles, sentence boundaries, and term-by-document structure.

Identifying Text Patterns Using NLP

Author LITI rules (CLASSIFIER, CONCEPT, CONCEPT_RULE, REGEX, SEQUENCE, PREDICATE_RULE) with SENT and DIST_n predicates; tune topic discovery via SVD with TF-IDF and log-entropy weights; build boolean and supervised category models; score sentiment with PROC SENTIMENT, polarity classes, negation handling, and feature-level sentiment; train RNN/LSTM text classifiers with embeddings; score with PROC TMSCORE.

Identifying Image Patterns Using Computer Vision

Build CNN topologies via deepLearn.buildModel and addLayer (convolution, padding, pooling, batch norm, dropout, fully connected, softmax); train with dlTrain (ADAM, SGD, learning rate, weight decay, augmentation); score with dlScore; apply transfer learning with ResNet-50, VGG-16, and MobileNet; design object detection with YOLO and Faster R-CNN, anchor boxes, NMS, IoU, and mAP; deploy via the astore action set.