Semi-Supervised & Self-Supervised Learning Quiz

1. Which semi-supervised learning method assumes that decision boundaries pass through low-density regions?

Transductive SVM

K-means clustering

Random forest

Naive Bayes classifier

2. True or False: Self-supervised learning always outperforms supervised learning on all tasks.

True

False

3. Which self-supervised method predicts masked tokens in sequences, similar to BERT?

Masked language modeling

Next sentence prediction

Token classification

Sequence alignment

Submit

5. What is the assumption underlying graph-based semi-supervised methods?

Similar data points should have similar labels

All data must be represented as vectors

Labels are uniformly distributed across classes

Features are always independent

6. True or False: Semi-supervised learning requires an equal balance of labeled and unlabeled data.

True

False

Submit

8. Which approach combines labeled and unlabeled data by optimizing both supervised and unsupervised objectives?

Mixed supervision learning

Entropy minimization

Temporal ensembling

Adversarial training

9. What is the role of data augmentation in self-supervised learning?

Creating different views of the same sample for contrastive learning

Increasing dataset size artificially

Reducing model overfitting only

Normalizing feature distributions

Submit

11. What is the primary advantage of semi-supervised learning over purely supervised learning?

It requires less labeled data while utilizing unlabeled data

It eliminates the need for any labeled data

It always produces better accuracy than supervised methods

It reduces computational complexity significantly

Submit

13. What is pseudo-labeling in semi-supervised learning?

Assigning labels to unlabeled data based on model predictions

Using only labeled data for training

Creating synthetic labels randomly

Removing unreliable labels from datasets

Submit

15. Which of the following is an example of a pretext task for self-supervised vision learning?

Predicting rotations of images

Classifying objects with human labels

Measuring pixel intensity

Applying Gaussian blur filters

Submit

17. What is contrastive learning in self-supervised learning?

Learning by contrasting similar and dissimilar data points

Comparing supervised versus unsupervised methods

Using contrast enhancement in image preprocessing

Balancing positive and negative samples equally

Submit

19. Which technique is commonly used in semi-supervised learning to propagate labels from labeled to unlabeled data?

Clustering and label propagation

Random forest ensemble

Principal component analysis

Support vector machines only

20. In self-supervised learning, what is a pretext task?

A task used to generate pseudo-labels from unlabeled data

A supervised task requiring human annotation

A task that requires external databases

A task that measures model generalization

DataAI Semi Supervised and Self Supervised Learning Quiz

1. Which semi-supervised learning method assumes that decision boundaries pass through low-density regions?

2.

What first name or nickname would you like us to use?

2. True or False: Self-supervised learning always outperforms supervised learning on all tasks.

3. Which self-supervised method predicts masked tokens in sequences, similar to BERT?

4. In BYOL (Bootstrap Your Own Latent), the learning mechanism does not require ____.

5. What is the assumption underlying graph-based semi-supervised methods?

6. True or False: Semi-supervised learning requires an equal balance of labeled and unlabeled data.

7. Self-supervised learning can be used as a pretraining strategy before ____ on downstream tasks.

8. Which approach combines labeled and unlabeled data by optimizing both supervised and unsupervised objectives?

9. What is the role of data augmentation in self-supervised learning?

10. In self-supervised learning, momentum contrast (MoCo) maintains a ____ of negative samples.

11. What is the primary advantage of semi-supervised learning over purely supervised learning?

12. Self-supervised learning approaches like SimCLR use ____ to learn representations.

13. What is pseudo-labeling in semi-supervised learning?

14. In semi-supervised learning, the consistency regularization principle enforces that ____.

15. Which of the following is an example of a pretext task for self-supervised vision learning?

16. Co-training is a semi-supervised method that uses ____ different views of data.

17. What is contrastive learning in self-supervised learning?

18. Self-supervised learning generates supervision signals from ____.

19. Which technique is commonly used in semi-supervised learning to propagate labels from labeled to unlabeled data?

20. In self-supervised learning, what is a pretext task?