CS362 - Midterm

Question 1

Artificial Intelligence (AI)

Answer 1

A non-human software/model capable of doing complex tasks like interpreting text or diagnosing illnesses.

Week 2: Introduction to Artificial Intelligence

Question 2

Machine Learning (ML)

Answer 2

A subset of AI where systems train on data to make predictions.

Week 2: Introduction to Artificial Intelligence

Question 3

Deep Learning (DL)

Answer 3

A type of machine learning using neural networks with multiple hidden layers.

Week 2: Introduction to Artificial Intelligence

Question 4

Weak AI (Narrow AI)

Answer 4

AI that performs a specific task only (e.g., facial recognition).

Week 2: Introduction to Artificial Intelligence

Question 5

Strong AI

Answer 5

Hypothetical AI that can perform any intellectual task a human can.

Week 2: Introduction to Artificial Intelligence

Question 6

Artificial General Intelligence (AGI)

Answer 6

AI with human-level intelligence across tasks.

Week 2: Introduction to Artificial Intelligence

Question 7

Supervised Learning

Answer 7

Learning from labeled data to map inputs to outputs.

Week 2: Introduction to Artificial Intelligence

Question 8

Unsupervised Learning

Answer 8

Learning from unlabeled data by finding patterns or groupings (e.g., clustering).

Week 2: Introduction to Artificial Intelligence

Question 9

Semi-supervised Learning

Answer 9

Training on data with some labeled and some unlabeled examples.

Week 3: Machine Learning

Question 10

Self-supervised Learning

Answer 10

Learning by generating labels from the data itself (e.g., denoising).

Week 3: Machine Learning

Question 11

Probabilistic Modelling

Answer 11

Statistical modeling using probability to predict outcomes.

Week 3: Machine Learning

Question 12

Decision Tree

Answer 12

A supervised model using decision rules in a tree structure.

Week 3: Machine Learning

Question 13

Random Forest

Answer 13

An ensemble of decision trees using bagging.

Week 3: Machine Learning

Question 14

Gradient Boosting Machine (GBM)

Answer 14

Builds models sequentially to correct previous errors.

Week 3: Machine Learning

Question 15

Hyperparameter

Answer 15

A setting used to tune a learning algorithm (not learned from data).

Week 3: Machine Learning

Question 16

Cross-validation

Answer 16

A method to prevent overfitting by testing on multiple data splits.

Week 3: Machine Learning

Question 17

Overfitting

Answer 17

When a model performs well on training data but poorly on new data.

Week 3: Machine Learning

Question 18

Linear Regression

Answer 18

Models the relationship between inputs and a continuous output.

Week 3: Machine Learning

Question 19

Binary Classification

Answer 19

Classifies data into one of two categories.

Week 3: Machine Learning

Question 20

Multi-class Classification

Answer 20

Classifies data into more than two categories.

Week 3: Machine Learning

Question 21

Ensemble Learning

Answer 21

Combines multiple models to improve performance.

Week 3: Machine Learning

Question 22

Tensors

Answer 22

Multi-dimensional arrays used in deep learning to represent data.

Week 4: Deep Learning

Question 23

Feedforward Neural Network

Answer 23

A basic neural network where data flows in one direction.

Week 4: Deep Learning

Question 24

Computation Graph

Answer 24

A directed graph representing operations and data flow in a model.

Week 4: Deep Learning

Question 25

Convolutional Neural Network (CNN)

Answer 25

A neural network specialized in visual data. Week 4: Deep Learning

Question 26

Recurrent Neural Network (RNN)

Answer 26

A neural network for sequential data (e.g., text, time series). Week 4: Deep Learning

Question 27

Backpropagation

Answer 27

Algorithm for adjusting weights in neural networks based on error. Week 4: Deep Learning

Question 28

Gradient Descent

Answer 28

Optimization algorithm to minimize the loss function. Week 4: Deep Learning

Question 29

Generalization

Answer 29

A model’s ability to perform well on unseen data. Week 4: Deep Learning

Question 30

Transfer Learning

Answer 30

Using a pre-trained model for a different but related task. Week 4: Deep Learning

Question 31

Computer Vision

Answer 31

Enabling machines to understand and interpret visual inputs. Week 5: Deep Learning for Computer Vision

Question 32

Pixel

Answer 32

The smallest unit of a digital image representing one color. Week 5: Deep Learning for Computer Vision

Question 33

Image Classification

Answer 33

Assigning a label to an image. Week 5: Deep Learning for Computer Vision

Question 34

Object Detection

Answer 34

Identifying and locating objects in an image. Week 5: Deep Learning for Computer Vision

Question 35

Image Segmentation

Answer 35

Dividing an image into parts representing different objects. Week 5: Deep Learning for Computer Vision

Question 36

Image Generation

Answer 36

Creating new images using models like Generative adversarial networks (GAN). Week 5: Deep Learning for Computer Vision

Question 37

Data Augmentation

Answer 37

Increasing training data using transformations. Week 5: Deep Learning for Computer Vision

Question 38

Pose Estimation

Answer 38

Estimating the position of human body parts in images. Week 5: Deep Learning for Computer Vision