ITM 209 Artificial Intelligence

What is Artificial intelligence

-technology, not specific to software or hardware -requires software and a database to communicate with it, and hardware to run it

How does machine learning work

data is gathered through human input,

Primary role of AI

interpret, analyze and respond to data (input given ie. typing or speaking)

Deep learning

is the most specific form of AI, subcategory of machine learning, allows us to model the human brain and teach the AI context of the situation, requires a lot of computer power

AI moves from

literal language to figurative language, and denotation to connotation

What are Neural Networks

-a computer system that models the human brain through the use of artificial neurons -utilizes AI, ML, DL NLP and NLU

Confidence scores

-range between 0-1 (1 being most accurate) -tells how accurate match need to be

machine learning purpose

-reduce the time needed to train an AI model -allows AI to get more intelligent over time -increase the size of the database

Literal Language

A form of language in which writers and speakers mean exactly what their words denote.

Natural Language Processing (NLP)

A technology that converts human language (structured or unstructured) into data that can be translated then manipulated by computer systems; branch of artificial intelligence

What components are needed for NLP

AI, ML, DL

Figurative language

Writing or speech that is not intended to carry literal meaning and is usually meant to be imaginative and vivid.

Connotation

an idea or feeling that a word invokes in addition to its literal or primary meaning.

Parameters

control and limit the actions of the AI

Current use for deep learning

predicting earthquakes, data augmentation, brain cancer detection, stock market predictions, crime pattern analysis

When is NLP needed/used

required for speech recognition, transcription, translation, allows for normal conversations, anytime "emotion" is expressed by the AI

Machine Learning

the extraction of knowledge from data based on algorithms created from training data

Denotation

the literal meaning of a word

What are the four questions NLU tries to answer?

what is being said, how it is said, why they said it, and what are the implications