ITCT101: AI, ML, and Data Science - Adoption & Types
39 Questions
0 Views

Choose a study mode

Play Quiz
Study Flashcards
Spaced Repetition
Chat to Lesson

Podcast

Play an AI-generated podcast conversation about this lesson

Questions and Answers

According to Tom Mitchell's definition, what are the three key elements in machine learning?

  • Tasks, performance, and data
  • Tasks, experience, and programming
  • Experience, data, and tasks
  • Experience, performance, and tasks (correct)

If a machine learning program is designed to predict the 'right' number from pairs of numbers, and it predicts (5, 6) when the correct answer is (5, 10), what is the performance score according to the provided example?

  • 0% (correct)
  • 60%
  • 100%
  • 50%

In the machine learning process, what is the primary purpose of the 'training dataset'?

  • To provide data for the learning algorithm to create a model. (correct)
  • To reduce the dimensionality of the data.
  • To detect anomalies in the data.
  • To evaluate the model's final performance.

Which of the following is NOT one of the four common learning types in machine learning?

<p>Descriptive Learning (A)</p> Signup and view all the answers

Which of the following machine learning tasks would be classified as 'unsupervised learning'?

<p>Identifying customer segments based on purchasing behavior. (C)</p> Signup and view all the answers

Which of the following is an example of dimensionality reduction?

<p>Reducing the number of variables in a dataset while preserving its variance. (C)</p> Signup and view all the answers

Why might smaller machine learning models be more expensive to operate, despite their reduced size?

<p>Smaller models may require more complex training regimes or feature engineering, indirectly raising expenses. (C)</p> Signup and view all the answers

In reinforcement learning, what is the primary objective of the agent?

<p>To learn a policy that maximizes cumulative rewards through interaction with an environment. (B)</p> Signup and view all the answers

Consider a scenario where you have a limited amount of labeled data and a large amount of unlabeled data. Which learning type would be MOST appropriate?

<p>Semi-supervised Learning (C)</p> Signup and view all the answers

A self-driving car is learning to navigate a complex environment. It receives feedback in the form of rewards (reaching the destination) and penalties (collisions). Over time, it adjusts its driving strategy to maximize its rewards. Which type of machine learning is best suited for this task?

<p>Reinforcement Learning (D)</p> Signup and view all the answers

What is the primary distinction between Artificial Intelligence (AI) and Machine Learning (ML)?

<p>AI focuses on tasks, while ML learns to perform tasks from data. (B)</p> Signup and view all the answers

What is the role of a 'training dataset' in machine learning?

<p>To provide data for the model to learn the relationship between inputs and labels. (B)</p> Signup and view all the answers

In the context of machine learning, what does a 'model' capture?

<p>The relationship between input data and their corresponding labels. (D)</p> Signup and view all the answers

What is the purpose of the 'inferencing' stage in machine learning?

<p>To predict outputs for new, unseen inputs using the trained model. (B)</p> Signup and view all the answers

The rise of which algorithm significantly contributed to the development of deep learning in 1986?

<p>Backpropagation. (C)</p> Signup and view all the answers

What is a defining characteristic of deep learning models in relation to data?

<p>They scale well with the amount of data. (C)</p> Signup and view all the answers

Which of the following best describes the relationship between Deep Learning, Machine Learning, and Artificial Intelligence?

<p>Deep Learning ⊆ ML ⊆ AI (C)</p> Signup and view all the answers

Consider a scenario where a model is deployed and consistently makes inaccurate predictions on a specific subset of new inputs. What could be the underlying problem?

<p>The training dataset did not adequately represent the characteristics of this subset of inputs. (D)</p> Signup and view all the answers

A machine learning engineer aims to build a model that can accurately predict stock prices. They have access to a vast dataset of historical stock prices and relevant economic indicators. Which of the following approaches would be MOST suitable?

<p>Implement a deep learning model with recurrent neural networks (RNNs) to capture temporal dependencies in the data. (B)</p> Signup and view all the answers

A researcher observes that increasing the size of a deep learning model leads to diminishing returns in performance. After a certain point, adding more layers or parameters yields only marginal improvements. What could be the underlying reason?

<p>The available dataset is not large enough to support the model's capacity, or the model is not complex enough to learn new patterns. (D)</p> Signup and view all the answers

Which of the following best describes Artificial Narrow Intelligence (ANI)?

<p>AI systems designed to perform excellently at a specific, narrow task. (A)</p> Signup and view all the answers

Which of the following is an example of a task typically associated with Artificial Narrow Intelligence?

<p>Providing personalized movie recommendations based on user preferences. (A)</p> Signup and view all the answers

What is the primary distinction between Artificial General Intelligence (AGI) and Artificial Narrow Intelligence (ANI)?

<p>AGI can perform a wide range of intellectual tasks, while ANI is limited to specific applications. (C)</p> Signup and view all the answers

Which of the following best describes an example of Machine Learning?

<p>A spam filter that improves its accuracy over time by learning from user feedback. (C)</p> Signup and view all the answers

Which of the following statements is most accurate regarding the current state of Artificial General Intelligence (AGI)?

<p>AGI is still a theoretical concept, and a true AGI system does not yet exist. (A)</p> Signup and view all the answers

What is the significance of the year 1956 in the context of Artificial Intelligence?

<p>The term Artificial Intelligence was coined at the Dartmouth Workshop. (A)</p> Signup and view all the answers

Which of the following represents the chronological order of these milestones?

<p>Linear Regression -&gt; Reasoning with uncertainty: Bayesian networks -&gt; Support vector machines (B)</p> Signup and view all the answers

In what year was the term 'Artificial Intelligence' coined?

<p>1956 (C)</p> Signup and view all the answers

Consider the following scenario: A company wants to implement an AI system to automate customer service inquiries. Which type of AI would be most suitable for this task?

<p>Artificial Narrow Intelligence (ANI) (A)</p> Signup and view all the answers

What distinguishes Artificial General Intelligence (AGI) from Artificial Narrow Intelligence (ANI)?

<p>AGI can perform any intellectual task that a human being can, while ANI is limited to specific tasks. (A)</p> Signup and view all the answers

Let's say you are trying to classify images, and need the highest possible accuracy. You have a massive dataset, but limited computational resources. Which algorithm would be the LEAST suitable choice?

<p>Linear Regression (B)</p> Signup and view all the answers

A team of AI researchers is attempting to create an AGI. They have developed a system that can excel at tasks it was not explicitly trained for, by leveraging previously learned concepts and skills from disparate domains. However, the crucial bottleneck preventing them from achieving true AGI is that the system still struggles with:

<p>Adapting and reasoning in truly novel, unpredictable real-world situations, demonstrating genuine common sense. (C)</p> Signup and view all the answers

Which of the following is an example of Artificial Narrow Intelligence (ANI)?

<p>AlphaGo playing the game of Go. (D)</p> Signup and view all the answers

According to the McKinsey Global Survey referenced, what can be inferred regarding AI adoption rates?

<p>AI adoption varies, with some industries and experts leveraging AI more than others. (C)</p> Signup and view all the answers

Consider an AI system designed to diagnose medical conditions from patient data. If the system performs exceptionally well in controlled clinical trials but struggles with real-world patient data due to variations in data quality and patient demographics, what limitation of AI is most evident?

<p>Dependence on high-quality, representative training data. (D)</p> Signup and view all the answers

Imagine an AI powered customer service chatbot. Which scenario would likely require a transition to a human agent?

<p>Handling a customer complaint involving complex emotional factors and requiring nuanced understanding. (D)</p> Signup and view all the answers

AlphaGo's victory over a Go grandmaster in 2016 demonstrated a significant milestone in AI. Which of the following best describes the primary technological advancement that enabled this achievement?

<p>Advancements in deep learning and reinforcement learning. (D)</p> Signup and view all the answers

Consider a scenario where an AI system is used to predict criminal recidivism rates. If the system is trained on historical data that reflects existing biases in law enforcement practices, what is the most likely ethical concern that will arise?

<p>The AI will perpetuate and amplify existing biases, leading to unfair or discriminatory outcomes. (A)</p> Signup and view all the answers

An AI company is developing a new product that utilizes facial recognition technology. To ensure responsible and ethical AI development, which action would offer the most significant impact in mitigating potential harms related to bias and privacy?

<p>Hiring a diverse team of AI developers and ethicists to actively identify and address potential biases in the data and algorithms. (D)</p> Signup and view all the answers

Flashcards

Artificial Intelligence (AI)

Machines capable of performing tasks in an intelligent manner.

AI's Founding Year

The year the term 'Artificial Intelligence' was coined at Dartmouth College.

Artificial Narrow Intelligence (ANI)

AI focused on performing a specific task exceptionally well.

Artificial General Intelligence (AGI)

A category of AI that can perform any intellectual task that a human being can.

Signup and view all the flashcards

AlphaGo

An AI program designed to play and beat human experts at the game of Go.

Signup and view all the flashcards

AI Adoption Rate

Indicates that many industries are incorporating AI into their operations at an increasing rate.

Signup and view all the flashcards

Weak AI

AI systems excelling at a particular task.

Signup and view all the flashcards

Intelligence

The ability of machines to reason, discover meaning, generalize and learn from past experience

Signup and view all the flashcards

Artifical

The simulation of human intelligence processes by machines, especially computer systems

Signup and view all the flashcards

Tasks of ANI

AI systems performing well on a specific task.

Signup and view all the flashcards

Artificial Narrow Intelligence

AI focused on performing a specific task well.

Signup and view all the flashcards

Artificial General Intelligence

AI that can perform many intellectual tasks that a human can.

Signup and view all the flashcards

Machine Learning

Machines learn to perform tasks from data.

Signup and view all the flashcards

Linear Regression

Predicting a continuous outcome based on input variables.

Signup and view all the flashcards

Linear Classifier

Separating data into categories using a linear boundary.

Signup and view all the flashcards

Bayesian Networks

Reasoning and decision-making under conditions of uncertainty.

Signup and view all the flashcards

Support Vector Machines

Finding the optimal boundary to separate different classes of data.

Signup and view all the flashcards

Training Dataset

Data used to train a machine learning model, often including inputs and their corresponding labels.

Signup and view all the flashcards

Learning Algorithm

An algorithm that learns patterns from data.

Signup and view all the flashcards

Model

A representation of the relationship between inputs and outputs, learned from training data.

Signup and view all the flashcards

Machine Learning (ML)

A subset of AI focused on enabling machines to learn from data without explicit programming.

Signup and view all the flashcards

Inferencing

Using a trained model to generate outputs (predictions) for new, unseen inputs.

Signup and view all the flashcards

Prediction

The output generated by a model based on a given input.

Signup and view all the flashcards

Deep Learning (DL)

A subset of machine learning that uses artificial neural networks with multiple layers to analyze data.

Signup and view all the flashcards

Supervised Learning

Models that learn from data with the benefit of explicit labels or targets.

Signup and view all the flashcards

Large Models

Deep Learning models that often use a high number of artifical neural networks.

Signup and view all the flashcards

Machine Learning Definition

A computer improves performance (P) on tasks (T) with experience (E).

Signup and view all the flashcards

Machine Learning Model Creation

Training data + algorithm = Model

Signup and view all the flashcards

Semi-Supervised Learning

Combines labeled and unlabeled data for learning.

Signup and view all the flashcards

Reinforcement Learning

Learning through trial and error, receiving rewards or penalties.

Signup and view all the flashcards

Clustering

Grouping similar data points together.

Signup and view all the flashcards

Dimensionality Reduction

Reducing the number of variables while preserving important information.

Signup and view all the flashcards

Anomaly Detection

Identifying rare or unusual data points.

Signup and view all the flashcards

What is evaluated by the performance measure

Pairs of numbers, program predicts the correct number and performance is based on correct guesses

Signup and view all the flashcards

Signup and view all the flashcards

Signup and view all the flashcards

Study Notes

  • ITCT101 Computer Technologies, Module 2 covers AI, ML, and Data Science

AI Adoption Rate

  • McKinsey's 2024 Global Survey shows the adoption rate of AI in at least one business function
  • There is an increasing use of generative AI

What is Artificial Intelligence (AI)?

  • AI, coined in 1956 by John McCarthy and colleagues, refers to machines that perform tasks in an intelligent manner
  • There are two categories of AI: Artificial Narrow Intelligence and Artificial General Intelligence

Artificial Narrow Intelligence (Weak AI)

  • Artificial Narrow Intelligence (Weak AI) includes AI applications skilled at specific tasks
  • These include AlphaGo, which won against a Go grandmaster in 2016, diagnosis and pre-screening tools, movie recommendations, spam mail detection and image restoration

Artificial General Intelligence (Strong AI)

  • Artificial General Intelligence (Strong AI) includes AI applications capable of doing mant intellectual tasks that a human can do
  • This type of AI does not yet exist, with examples that are often portrayed in movies

Machine Learning (ML)

  • Machine Learning, developed in the 1990s, is a field of AI where machines learn to perform tasks from data
  • The process involves training a learning algorithm on a training dataset to create a model
  • The model is then used to predict an output for a new, often unseen, input

Deep Learning

  • Deep Learning, which emerged in the 2010s, is a subset of machine learning techniques based on artificial neural networks
  • Deep learning models scale well with large amounts of data

Data Science

  • Data Science includes using data, learning from it, and creating useful models to understand trends, find insights, and make informed decisions

Machine Learning - Definition of Learning

  • According to Tom Mitchell, a computer program learns from experience E with respect to tasks T and a performance measure P
  • Performance at tasks in T, measured by P, improves with experience E

Example of Machine Learning

  • A machine learning task is to predict the correct number in pairs
  • The machine's learning performance is measured by the percentage of correct guesses
  • Over time, the computer program is fed data to learn from: (1, 2), (2, 4), (3, 6), (4, 8)

Machine Learning (ML) Process

  • The Machine Learning Process is as follows: A Learning Algorithm learns from a Training Dataset so generate a Model
  • A Testing Dataset is used to compare the Models performance with its Evaluation

Common Learning Types

  • Supervised Learning: Learning a function that maps an input x, to an output Å·, using training samples of inputs and labels
    • Involves Regression and Classification
  • Unsupervised Learning: Learning a function that maps an input x, to an output Å·, using training samples of inputs
    • Involves Clustering, Dimensionality reduction, and Anomaly detection
  • Semi-supervised Learning: Learning a function that maps an input x, to an output Å·, using training
    • Involves labeled and unlabeled samples of inputs
  • Reinforcement Learning: Learning a function that maps an input x, to an output Å·, using training samples based on interaction (observation and action) and reward

Studying That Suits You

Use AI to generate personalized quizzes and flashcards to suit your learning preferences.

Quiz Team

Related Documents

Description

Module 2 of ITCT101 explores AI, ML, and Data Science. The 2024 McKinsey survey shows increasing AI adoption. Explores key concepts: Artificial Narrow Intelligence (Weak AI) and Artificial General Intelligence (Strong AI).

More Like This

Use Quizgecko on...
Browser
Open
Browser
Browser