wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Meta-Learning, Few-Shot & Zero-Shot Learning & Graph

Total questions: 20

Worksheet time: 10mins

Name
Class
Date
1.

Meta-learning means

a)

Learning faster models

b)

Learning how to learn

c)

Learning without data

d)

Learning deeper networks

2.

Few-shot learning refers to

a)

Many labeled samples

b)

Learning from very few examples

c)

Learning without labels

d)

Learning once

3.

Zero-shot learning means

a)

No training

b)

No testing

c)

Recognizing unseen classes

d)

No model

4.

Meta-learning operates across

a)

Samples

b)

Tasks

c)

Classes

d)

Features

5.

A task in meta-learning contains

a)

One sample

b)

Support set and query set

c)

Only test data

d)

Only training data

6.

Query set is used for

a)

Learning representations

b)

Updating meta-parameters

c)

Evaluating task performance

d)

Data augmentation

7.

Prototypical Networks are

a)

Gradient-based

b)

Metric-based

c)

Rule-based

d)

Probabilistic

8.

Prototype represents

a)

Random sample

b)

Mean of class embeddings

c)

Maximum value

d)

Weight vector

9.

Distance metric commonly used is

a)

Manhattan

b)

Cosine

c)

Euclidean

d)

Hamming

10.

Meta-training differs from normal training because

a)

No loss

b)

Trains across tasks

c)

Uses test data

d)

No labels

11.

Zero-shot learning relies on

a)

Attribute information

b)

Random guessing

c)

Large batch sizes

d)

CNN depth

12.

Few-shot learning is common in

a)

NLP & Vision

b)

Databases

c)

Networking

d)

Operating systems

13.

One-shot learning means

a)

One task

b)

One example per class

c)

One epoch

d)

One model

14.

Meta-learning improves

a)

Memory usage

b)

Generalization to new tasks

c)

Training loss only

d)

Batch size

15.

Graph data consists of

a)

Rows and columns

b)

Nodes and edges

c)

Pixels

d)

Tokens

16.

Graphs are useful for

a)

Images only

b)

Sequential data

c)

Relational data

d)

Tabular data

17.

Directed graph means

a)

No edges

b)

Bidirectional edges

c)

Edges have direction

d)

No nodes

18.

Undirected graphs have

a)

One-way edges

b)

Two-way edges

c)

No attributes

d)

No nodes

19.

Graph Machine Learning learns from

a)

Independent samples

b)

Connected data

c)

Images

d)

Text

20.

Examples of graph data include

a)

Molecules

b)

Road networks

c)

Knowledge graphs

d)

All of the above