Machine Learning Module 2

Question 1

What the are 2 types of Supervised Learning?

Answer 1

Classification and Regression

Question 2

What is Classification?

Answer 2

Classification is the process of predicting discrete class labels or categories. i.e. Will I pass or fail my biology exam?

Question 3

What is Regression?

Answer 3

Regression is the relationship between a dependent and independent variable. Regression is the process of predicting continous values.

Question 4

Difference between Classification and Regression?

Answer 4

Classification works with classes while regression is mapped to a continous variable

Question 5

What are some commong algorithms for regression?

Answer 5

Random forests, gradient boosting, support vector regression(SVR) and neural networks

Question 6

What is Unsupervised learning?

Answer 6

The model is not supervised and allowed to work independently to discover patterns and structures in the data that is not visible to the human eye. This means the model trains on the data set and draws conclusions on its own.

Question 7

Types of Unsupervised learning?

Answer 7

Dimension reduction, Density estimation, Market basket analysis and Clustering.

Question 8

What is Dimension Reduction?

Answer 8

Also known as feature selection, seeks to reduce redundant features to make the classification easier.

Question 9

What is Density Estimation?

Answer 9

Straight forward concept to explore data and find a structure

Question 10

What is Market Basket Analysis?

Answer 10

Modelling technique based on the theory that if you buy a group of items you are more likely to buy another group of items

Question 11

What is Clustering?

Answer 11

Most popular unsupervised learning technique and is the Grouping of Data Points or Objects that are similar by the characteristics of the data.

Question 12

What is Clustering used for?

Answer 12

To detect Structure, to Summarize and to detect anomalies within the dataset.

Question 13

What is the main difference between Supervised Learning and Unsupervised Learning?

Answer 13

Supervised learning deals with ‘Labelled’ data and Unsupervised learning deals with ‘Un-Labelled’ data.

Question 14

What is a classifier?

Answer 14

A machine learning algorithm used to solve the classification problem.

Question 15

What is a feature?

Answer 15

In independent variable that is used as an input into the model

Question 16

What is evaluation?

Answer 16

Assessing how well a model has performed.

Question 17

What are 2 types of Classification Algorithms?

Answer 17

Lazy learner and Eager learner

Question 18

Describe the Eager learner classification model

Answer 18

Spends a lot of time training and generalizing the model

Question 19

Describe the Lazy learner classification model

Answer 19

Doesn’t have a training phase per say and waits to have a test data set before making predictions. So it takes longer to predict as it doesn’t generalize the model.

Question 20

What is an example of a lazy learner classification algorithm?

Answer 20

K-Nearest Neighbours or KNN.

Question 21

How does KNN work?

Answer 21

KNN classifies the unknown data points by finding the most common classes in the K nearest examples. Then finds the closest match to the test data.

Question 22

What is an example of an eager learning classification algorithm?

Answer 22

Logistic regression that is used to predict the probability of a class based on a graph plot

Question 23

What is another example of an eager learning classification algorithm?

Answer 23

Decision trees, that are a tree like algorithm that use an ‘If’ ‘Then’ rule or set of rules. So it classifies if you pass or fail based on rules.

Question 24

Other classificaiton algorithms

Answer 24

Support vector machines, discriminate analysis, naive Bayes, and neural networks.

Question 25

What is a confusion matrix?

Answer 25

A confusion matrix measures the performance and accuracy of a classification problem

Question 26

How is the mean squared error helpful?

Answer 26

The mean squared error is helpful for evalutating regression models

Question 27

What is the R-squared value?

Answer 27

The R-squared value represents the amount of variance in a dependent variable that can be explained by an independent variable

Question 28

What can be used to evauluate classificaiton models?

Answer 28

A confusion matrix, precision and recall

Question 29

What does train/test split mean?

Answer 29

This is the process of splitting data into ‘training’ and ‘testing’. Some data is used to train the machine learning model and the rest is used to test it.

Question 30

What is Deep Learning?

Answer 30

Deep Learning is a subset of machine learning and has 3 layers, Input, Hidden and Output. Deep learning is loosely based on the human brain.

Question 31

What is Artificial Neural Network?

Answer 31

It is a group of connected nodes called neurons. ANN loosely models the human brain

Question 32

What is a Neuron?

Answer 32

A neuron is a computation unit that can be expressed as a weighted sum of inputs. Each neuron has an activation function that predicts the value of the output.

Question 33

What is a perceptron?

Answer 33

The coinfiguration of all inputs and their weights is a perceptron. It also contains a summation function which is a total summation of the inputs and their associated weights.

Question 34

What is a Single-Layer neural network?

Answer 34

A neural network with one neuron.

Question 35

What is a ‘Multi-Layer Perceptron?

Answer 35

A neural network with an input layer, one processing layer(or hidden layer) and an output layer

Question 36

What is a Deep Neural Network?

Answer 36

An artificial neural network that has two or more hidden layers or two or more processin layers.

Question 37

What is Reinforcement Learning?

Answer 37

Reinforcement learning is a machine learning method. Reinforcement learning works like a reward system.

Question 38

What are the concepts of Reinforcement Learning?

Answer 38

The most common concepts in reinforcement learning are the agent, environment, reward, action, state, and policy.

Question 39

What are the types of Reinforcement Learning?

Answer 39

The two types of reinforcement learning are positive and negative reinforcement.

Question 40

What industries use reinforcement learning?

Answer 40

Reinforcement learning is used in various industries like games, fleet management, and manufacturing.

Question 41

In reinforcement learning what is a Policy?

Answer 41

A policy defines how the agent will act when in a specific state, it maps the states in the environment to a single action or probability for each action. A policy can be represented as a lookup table, a function, an algorithm, or by calculating the probability of performing an action given a state.

Question 42

What are the 3 algorithms for reinforcement learning?

Answer 42

Model based, Policy based and Value based

Question 43

What is the Model based algorithm for reinforcement learning?

Answer 43

Here, you create a virtual model to help the agent learn in specific environments

Question 44

What is the Policy based algorithm for reinforcement learning?

Answer 44

You develop a strategy that helps to gain maximum rewards in the future through possible actions performed in each state.

Question 45

What is the Value based algorithm for reinforcement learning?

Answer 45

The agent’s goal here is to maximize a defined value function.

Question 46

What are value functions?

Answer 46

Value functions are estimates for actions and states resulting from interacting with the environment.

Question 47

What is Generative AI

Answer 47

Generative Artificial Intelligence, or GenAI, is an AI technique capable of creating new and unique data, ranging from images and music to text and entire virtual worlds.

Question 48

How does Generative AI work?

Answer 48

Generative AI models use deep learning techniques and rely on vast datasets to generate entirely new data with various applications.