Overview

Teaching: 5 min
Exercises: 0 min

Questions

• What is machine learning? How is it different from traditional algorithms?

Objectives

• Get a basic understanding of machine learning.

What is Machine Learning (ML)?

Short answer: letting the computer learn rules from data.

In a typical scenario, we have a set of observed data consisting of:

• A set of input data (e.g., satellite image color, measurement time of year, location)
• Outcome measurements (e.g., landcover type)

Using this data, we build a prediction model, or learner, that is able to predict the outcome for new input data (note that this typical scenario is only one type of machine learning, called supervised learning; machine learning algorithms do not necessarily need outcome measurements, in which case unsupervised learning can be used to directly find patterns in the input data. We will talk more about these different types in the following pages).

Learning itself is the act of gradually improving performance on a task without being explicitly programmed. This process mimics human neurological functions (e.g., teaching a child to speak): repetition with more data results in a better learner.

Figure: Illustration of machine learning.

Key Points

• machine learning, traditional algorithms

---