Artificial Intelligence - Overview

Can be defined as the ability of a computer to perform tasks that require human intelligence.

• Machine Learning

• Artificial Intelligence

• Unsupervised Learning

A difficulty with explicit programming for AI is . . .

• there are too many scenarios to specify the appropriate action in each case.

• it does not tell you which factors led to its decision

The kind of machine learning where instances must be sorted into categories:

• regression problem

• classification problem

• expert system

The kind of machine learning where a continuous number is predicted;

• regression problem

• classification problem

• expert system

Predicting the price of a house, according to the value of various features:

• regression problem

• classification problem

• expert system

Which one requires that features be specified by the programmer in advance?

• deep learning

• machine learning without neural networks

Deciding whether an email is spam or not:

• regression problem

• classification problem

• expert system

Deciding whether an object is, eg. a car or a truck:

• regression problem

• classification problem

• expert system

Classification problem with machine learning from labelled instances:

• supervised learning

• unsupervised learning

• reinforcement learning

Classification problem with machine learning from unlabelled instances, which are grouped in clusters that have close values for features:

• supervised learning

• unsupervised learning

• reinforcement learning