NLP Text Classification 2

Question 1

Which regularization technique is commonly applied in neural text classifiers to prevent overfitting by randomly deactivating neurons during training?

Accepted Answer

Dropout

Answer

Batch normalization

Answer

L2 weight decay

Answer

Gradient clipping

Question 2

What is the key advantage of fine-tuning a pre-trained language model (e.g., BERT) for text classification compared to training from scratch?

Accepted Answer

It leverages general language knowledge learned from large corpora, needing far less task-specific data

Answer

It requires more labeled data to achieve good performance

Answer

It eliminates the need for any labeled training data

Answer

It always produces shorter inference times than simpler models

Question 3

In text classification, what problem does 'class imbalance' refer to?

Accepted Answer

One class having significantly more examples than others in the training set

Answer

Having too many features relative to the number of samples

Answer

Using different loss functions for different classes

Answer

Misalignment between training and test vocabulary

Question 4

Which technique involves creating synthetic minority-class examples to address class imbalance in text classification?

Accepted Answer

SMOTE (Synthetic Minority Over-sampling Technique)

Answer

Subsampling

Answer

Pruning

Answer

Dimensionality reduction

Question 5

What is hierarchical text classification?

Accepted Answer

Organizing categories into a tree structure where documents are classified from broad to specific

Answer

Classifying text based on its reading difficulty level

Answer

Ranking documents by relevance rather than assigning a single category

Answer

Using multiple layers of neural networks for classification

Question 6

When using cross-validation for evaluating a text classifier, what does 'k-fold' mean?

Accepted Answer

The data is divided into k equal parts, each used once as the validation set while the rest train the model

Answer

The dataset is split into k classes for stratified sampling

Answer

The model is trained k times with different random seeds to test stability

Answer

k hyperparameters are tuned simultaneously using grid search

Question 7

Which loss function is most commonly used when training a neural network for multi-class text classification?

Accepted Answer

Categorical Cross-Entropy

Answer

Mean Squared Error (MSE)

Answer

Binary Cross-Entropy

Answer

Hinge Loss