What Is Machine Learning? Core Concepts Explained Simply

Before defining machine learning, it helps to understand the two words that compose it.

1. A machine is a device containing one or more parts that can convert energy, typically providing power for chemical, thermal, or electronic processes.

2. Learning is the ability to improve behavior based on experience.

What Is Machine Learning?

According to Tom Mitchell, machine learning is:

“A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E.”

In this definition:

1. Task T denotes what the machine is trying to improve, such as prediction, classification, or clustering.

2. Experience E can be training data or any input the machine can learn from.

3. Performance P is a measure of how well the machine performs the task.

Machine Learning Consists of Two Main Parts: Learner and Predictor

1. The learner receives input/experience and acquires new skills.

2. Background knowledge can also assist the learner.

3. With data input and background knowledge, the learner builds a knowledge model.

4. This model contains information learned from the input and experience.

5. The problem or task (e.g., prediction, classification) is then handed to the predictor.

6. Using the well‑trained model, the predictor attempts to generate a solution.

7. The solution can be improved by adding additional input/experience.

8. This cycle repeats continuously.

Difference Between Machine Learning and Standard Programs

In machine learning, you provide the computer with:

1. Input (experience)

2. Output (the result corresponding to the input)

The system produces a model/program as output, which can then be used to perform tasks.

In a standard program, you provide the computer with:

1. Input

2. Output (how to process the input)

The program then directly yields the output.

These are the basic concepts of machine learning.