2.1.2 AI Flashcards
What is Artificial intelligence
The simulation of intelligent behaviour by a computer
which enables it to make decisions without human
intervention
What are the 3 enablers of AI
The exponential growth in the speed of computers
Huang’s Law
Cloud computing
Describe Narrow AI
AI focuses on a specific, singular or limited task
Examples include image recognition, hyper
personalisation, chatbots, predictive text.
Trained on specific tasks by data scientists
Correlates questions or assignments to a specific data
set to accomplish a task.
No self-awareness, consciousness, ability to think
Describe General AI
- Seeks machines that can handle a range of cognitive
tasks with little oversight. - The ability to learn, generalise, apply knowledge and plan
for the future. - Must consistently pass the Turing test.
- Single, general intelligence that possesses common
sense and creativity and expresses emotions.
What is Machine Learning
Machine learning (ML) is a subset of Artificial Intelligence (AI) in terms of algorithms that can adapt without following explicit instructions.
What can Machine learning do?
can train a machine how to learn
Automates analytical model building
Is a branch of artificial intelligence based on the
idea that systems can learn from data, identify
patterns, and make decisions with minimal human
intervention
What is Supervised learning
You provide a data set that has pictures of apples. Then another data set that lets the model know that those are pictures of apples. Finally, a new set of data that only contains pictures of apples. At this point the system can recognise what the fruit is
and will remember it.
What is Unsupervised learning
You provide a collection of pictures of different kinds of fruit. The model analyses this to try and recognise any patterns. The machine categorises the picture into different types based on their similarities, and one of these categories would be mostly apples.
What is semi-supervised learning
You provide the machine with a data set of images and ask it to identify a fruit. The machine picks an image of an apple and tells you it is an orange. You feedback to the system that it is actually an apple. The machine remembers this and picks again.
You provide feedback once more. This continues whilst the machine learns from its
experiences and feedback.
What are the different types of AGI testing
The Turing test
Wozniak/Goertzel’s Coffee Test
Nilsson’s Employment Test
Goertzel’s Robot College Student Test
What is Wozniak/Goertzel’s Coffee test
As a part of the coffee test, the AI application has to go to a house and make coffee. The programme has to go to any kitchen and find the ingredients required and then perform the task of making a coffee.
What is the Turing test
A remote human interrogator, within a fixed time frame, must distinguish
between a computer and a human subject based on their replies to various questions posed by the interrogator
What is Nilsson’s employment
To pass this test, the AI program should be able to perform jobs that are performed by humans
What is Goertzel’s Robot College Student test
This test advocates an AI being enrolled in a college and getting a degree using the same resources as other students enrolled for the same degree.
What is meant by conditional probability
In probability theory, conditional probability is a measure of the probability of an event given that (by assumption, presumption, assertion, or evidence) another event has occurred.
What is meant by suggestion engines
A suggestion or recommendation engine is a system that suggests products, services, and information to users based on an analysis of data
What is meant by pattern recognition
Pattern recognition is the use of machine learning algorithms to identify patterns. It has been applied in various fields such as image analysis, computer vision, healthcare,
What are the 4 neural networks
Massive models
Massive Data
Trial & Error
Deep learning
What is meant by massive models
DeepMind has created a deep neural network with more than 100 billion parameters. These are the values in the network that get tweaked over and over again during training. The size of a model is measured by the number of parameters it has - the more parameters, the more data it can soak up.
What is meant by massive data
Neural networks can be thought of as a “brute force” technique, characterised by a lack of intelligence but a huge amount of data, because they start with a blank slate, and they hammer their way through to an accurate model
What is meant by trial and error
Training a neural network is a highly iterative process, or a trial and error process. The process is to turn an idea into code, experiment, and evaluate the outcome and make changes and repeat the process until you get the expected or otherwise satisfactory outcome.
What is meant by deep learning
Deep learning is a type of machine learning that uses multiple layers to progressively extract higher-level features from the raw input.
What are common uses of AI
translation engines
* healthcare
* cyber security
* transport
* entertainment
* education (to detect plagiarism)
* voice command (voice to text/smart data)
* social networking (auto tagging)
* shopping (search/recommendation)
* finance (fraud/credit)
* communication (filtering)
What is meant by SLAM
is the computational problem of constructing or updating a map of an unknown environment while simultaneously keeping track of an object’s location within it
What is meant by language recognition
Natural language processing is a subfield of computer science
concerned with the interactions between computers and human language.
What is meant by machine translation
Machine translation is the process of automatically translating content from
one language (the source) to another (the target) without any human input.
What is meant by anomaly detection
Anomaly detection (aka outlier analysis) is a step in data mining that
identifies data points, events, and/or observations that deviate from a dataset’s normal behaviour.
What is meant by image recognition
Image recognition is the ability of software to identify objects, places, people,
writing and actions in images.
What are the types of problems AI can solve
- Simultaneous Location and Mapping (SLAM)
- language recognition
- machine translation
- anomaly detection
- image recognition
