Objective 2.4 Apply Problem-Specific Transformations to Data Sets CERTIFIED-DATA-SCIENCE-PRACTITIONER Practice Questions (Page 5)

Questions 21–23

1. 21foundation · easy

   Which of the following is a characteristic of one-hot encoding for word vectorization?

   • AIt creates a binary vector with a single 1 and all other positions 0
   • BIt produces dense vectors with semantic meaning
   • CIt reduces the dimensionality of the vocabulary
   • DIt uses word frequency counts to weight the vector
   Reveal Solution Discussion Select an answer first
2. 22application · medium

   A medical imaging startup has a small labeled dataset of X-ray images (5,000 images) and wants to train a classifier to detect pneumonia. They have access to a large unlabeled dataset of general medical images. They want to use a pre-trained convolutional neural network (CNN) to extract features from the X-ray images before training a small classifier. Which approach should they use?

   • ATrain a Word2Vec model on the image metadata and use the resulting embeddings as features.
   • BUse one-hot encoding of the image file names as features.
   • CApply TF-IDF to the pixel intensity values of each image.
   • DUse a pre-trained CNN (e.g., ResNet) trained on ImageNet, remove the final classification layer, and use the output of the penultimate layer as feature vectors for the X-ray images.
   Reveal Solution Discussion Select an answer first
3. 23foundation · easy

   What is the primary purpose of word vectorization in NLP?

   • ATo reduce the size of a text corpus by removing duplicate words
   • BTo split text into individual words for further analysis
   • CTo convert text tokens into numerical vectors that machine learning models can process
   • DTo identify the grammatical role of each word in a sentence
   Reveal Solution Discussion Select an answer first