Unit-2: Advanced Concepts of Modeling in AI

Flashcards for reviewing advanced AI modeling concepts, including AI, ML, DL, supervised/unsupervised learning, neural networks, and more.

What is Artificial Intelligence (AI)?

Any technique that enables computers to mimic human intelligence.

What does an artificially intelligent machine work on to produce a desired output?

Algorithms and data.

What is Machine Learning (ML)?

Enables machines to improve at tasks with experience by learning from new data and considering exceptions.

What is Deep Learning (DL)?

Enables software to train itself to perform tasks using vast amounts of data and multiple machine learning algorithms.

Which is the umbrella term that holds both Deep Learning and Machine Learning?

Artificial Intelligence.

How does the lecture define Deep Learning in relation to Machine Learning?

A very specific learning approach and a subset of Machine Learning.

What is Object Classification in Machine Learning?

Identifying and labeling objects within an image or data point.

What is Anomaly Detection in Machine Learning?

Finding unexpected things hiding in data, such as a sudden spike in heart rate.

What is Object Identification in Deep Learning?

Identifying and labeling objects within an image using powerful algorithms to determine what's in a picture and categorize those things.

What is Digit Recognition in Deep Learning?

Training computers to identify handwritten digits (0-9) within images.

What is Data?

Information in any form.

What are columns of a table called?

Features.

Where do you find the features in a dataset?

The columns of the tables.

What are labels in a dataset?

Special features that depend on the context of the problem and we are trying to solve

What is labeled data?

Data to which some tag/label is attached.

What is unlabeled data?

The raw form of data to which no tag is attached.

What is a training data set?

A collection of examples given to the model to analyze and learn.

What is a testing data set used for?

Used to test the accuracy of the model.

What does Al Modeling refer to?

Developing algorithms/models which can be trained to get intelligent outputs.

What are the two general classifications of Al models?

Learning-based and Rule-based.

What is the Rule Based Approach in Al modeling?

AI modelling where the relationships or patterns in data are defined by the developer, and the machine follows those rules.

What is a learning-based approach?

The computer learns how to do something by looking at examples or getting feedback, finding patterns or rules on its own.

What are the three categories of machine learning based models?

Supervised, unsupervised, and reinforcement learning.

In reinforcement learning, what is a 'state'?

The character's current position on the game screen.

In reinforcement learning, what is an 'action'?

Moving left, right, or jumping.

In reinforcement learning, what is the 'environment'?

The game world, including obstacles and paths.

What is a key characteristic of a dataset used in a supervised learning model?

The dataset is labeled.

What is a key characteristic of a dataset used in an unsupervised learning model?

The dataset is unlabelled.

What are the key characteristics and uses of Supervised Learning?

Deals with labelled data and is useful in real-world problems like predicting prices.

What are the key characteristics and uses of Unsupervised Learning?

Deals with unlabelled data and is useful in finding unknown patterns within data.

What does Reinforcement Learning enable the computer to do?

Enables a computer to make a series of decisions that maximize a reward metric without human intervention.

What is ML model family discovering?

Determine relationships, learn by training, and discover new patterns.

What are the two types of Supervised Learning models?

Classification and Regression models.

What are the main features of a Classification Model?

Data is classified according to labels and works on discrete datasets.

What type of data do Regression Models work on, and what do they predict?

Models work on continuous data to predict a continuous value based on input variables.

What are the two sub-categories of Unsupervised Learning models?

Clustering and Association models.

What is Clustering?

Dividing data points into different groups or clusters based on their similarity.

What is the key difference between Clustering and Classification?

Classification uses predefined classes, while clustering finds similarities between objects.

What does Association Rule mining help to do?

Find interesting relationships between variables in a database.

What are the main categories and sub-categories of Machine Learning models?

Classification, Regression, Clustering, and Association.

What are the two main types of Deep Learning models?

ANN (Artificial Neural Networks) and CNN (Convolutional Neural Networks).

In deep learning, what is the machine training for?

To train itself around the data.

What are Artificial Neural Networks (ANN)?

Modeled on the human brain and nervous system; able to automatically extract features without programmer input.

What are Convolutional Neural Networks (CNN)?

Can take in an input image, assign importance to various aspects/objects in the image, and differentiate one from another.

What is a key advantage of neural networks?

They are able to extract data features automatically without needing the input of the programmer.

What are the three layers of a neural network?

Input layer, hidden layers, and output layer.

What are some real-world applications of neural networks?

Facial recognition, customer support chatbots, and vegetable price prediction.

Explain the ANN

It is like a simple brain where the data passes through neurons, gets processed, and results in a decision.

Explain the CNN

It breaks the image into smaller pieces and looks at the features in each part, which helps it understand the image better.

What is Artificial Intelligence (AI)?

Any technique that enables computers to mimic human intelligence.

An artificially intelligent machine works on what to produce a desired output?

Algorithms and data.

Machine Learning (ML) enable machines to improve at tasks with experience?

Enables machines to improve at tasks with experience by learning from new data and considering exceptions.

Deep Learning (DL) enable software to train itself to perform tasks using vast amounts of data?

Enables software to train itself to perform tasks using vast amounts of data and multiple machine learning algorithms.

Which is the umbrella term that holds both Deep Learning and Machine Learning?

Artificial Intelligence.

How does the lecture define Deep Learning in relation to Machine Learning?

A very specific learning approach and a subset of Machine Learning.

Object Classification in Machine Learning is ?

Identifying and labeling objects within an image or data point.

What is Anomaly Detection in Machine Learning?

Finding unexpected things hiding in data, such as a sudden spike in heart rate.

Object Identification in Deep Learning is ?

Identifying and labeling objects within an image using powerful algorithms to determine what's in a picture and categorize those things.

What is Digit Recognition in Deep Learning?

Training computers to identify handwritten digits (0-9) within images.

What is Data?

Information in any form.

What are columns of a table called?

Features.

Where do you find the features in a dataset?

The columns of the tables.

What are labels in a dataset?

Special features that depend on the context of the problem and we are trying to solve

What is labeled data?

Data to which some tag/label is attached.

What is unlabeled data?

The raw form of data to which no tag is attached.

What is a training data set?

A collection of examples given to the model to analyze and learn.

A testing data set is .

Used to test the accuracy of the model.

Al Modeling refers to .

Developing algorithms/models which can be trained to get intelligent outputs.

What are the two general classifications of Al models?

Learning-based and Rule-based.

What is the Rule Based Approach in Al modeling?

AI modelling where the relationships or patterns in data are defined by the developer, and the machine follows those rules.

What is a learning-based approach?

The computer learns how to do something by looking at examples or getting feedback, finding patterns or rules on its own.

The three categories of machine learning based models consist of supervised, , and reinforcement learning.

Supervised, unsupervised, and reinforcement learning.

In reinforcement learning, a 'state' is ?

The character's current position on the game screen.

Moving left, right, or jumping are all examples of what component of Reinforcement Learning (fill in the blank)?

Moving left, right, or jumping.

In reinforcement learning, what is the \'environment\'?

The game world, including obstacles and paths.

What is a key characteristic of a dataset used in a supervised learning model?

The dataset is labeled.

What is a key characteristic of a dataset used in an unsupervised learning model?

The dataset is unlabelled.

What are the key characteristics and uses of Supervised Learning?

Deals with labelled data and is useful in real-world problems like predicting prices.

What are the key characteristics and uses of Unsupervised Learning?

Deals with unlabelled data and is useful in finding unknown patterns within data.

What does Reinforcement Learning enable the computer to do?

Enables a computer to make a series of decisions that maximize a reward metric without human intervention.

ML model family discovering means .

Determine relationships, learn by training, and discover new patterns.

What are the two types of Supervised Learning models?

Classification and Regression models.

The main features of a Classification Model are .

Data is classified according to labels and works on discrete datasets.

What type of data do Regression Models work on, and what do they predict?

Models work on continuous data to predict a continuous value based on input variables.

What are the two sub-categories of Unsupervised Learning models?

Clustering and Association models.

Clustering is .

Dividing data points into different groups or clusters based on their similarity.

What is the key difference between Clustering and Classification?

Classification uses predefined classes, while clustering finds similarities between objects.

Association Rule mining helps to .

Find interesting relationships between variables in a database.

Classifications, Regressions, and Association are the primary categories and sub-categories of Machine Learning models.

Classification, Regression, Clustering, and Association.

What are the two main types of Deep Learning models?

ANN (Artificial Neural Networks) and CNN (Convolutional Neural Networks).

In deep learning, the machine is training ?

To train itself around the data.

Artificial Neural Networks (ANN) are .

Modeled on the human brain and nervous system; able to automatically extract features without programmer input.

Convolutional Neural Networks (CNN) .

Can take in an input image, assign importance to various aspects/objects in the image, and differentiate one from another.

A key advantage of neural networks is ?

They are able to extract data features automatically without needing the input of the programmer.

The three layers of a neural network are .

Input layer, hidden layers, and output layer.

Some real-world applications of neural networks include .

Facial recognition, customer support chatbots, and vegetable price prediction.

It is like a simple brain where the data passes through neurons