\(\renewcommand{\vec}[1]{\mathbf{#1}}\)

\(\newcommand{\vecw}{\vec{w}}\) \(\newcommand{\vecx}{\vec{x}}\) \(\newcommand{\vecu}{\vec{u}}\) \(\newcommand{\veca}{\vec{a}}\) \(\newcommand{\vecb}{\vec{b}}\)

\(\newcommand{\vecwi}{\vecw^{(i)}}\) \(\newcommand{\vecwip}{\vecw^{(i+1)}}\) \(\newcommand{\vecwim}{\vecw^{(i-1)}}\) \(\newcommand{\norm}[1]{\lVert #1 \rVert}\) \(\newcommand{\fhat}{\hat{f}}\)

Exploration 3.2: Different Settings of Machine Learning

Overview of Different Settings

There are many different types of learning settings:

supervised, unsupervised, and semi-supervised learning

Supervised Learning: A Closer Look

In supervised learning, the training examples are labeled, and there is some unknown function that generates the data (with labels). The job of machine learning is to recover this hidden function from the labeled data, or to find a good approximation of it. This approximate function is known as the prediction rule learned from data.

We differentiate two types of function which correspond to two subsettings of supervised learning, classification and regression:

classification examples: binary, multiclass, and structured classification

When Supervised Learning is Useful

Supervised learning is particularly useful in the following scenarios:

Videos