Binary Classification

What Does "Binary" Mean?

The word binary means there are exactly two possible choices.

The model must choose one of these two classes when it receives new data.

Simple Example — Pass or Fail

Suppose we want to predict whether a student will pass an exam based on how many hours they studied.

The model learns from these examples.

How Does Binary Classification Work?

The model first learns from training examples where the correct class is already known.

The important part is that the model has only two possible classes to choose from.

Example — Spam or Not Spam

Email spam detection is another common Binary Classification problem.

The model can look at information such as:

Binary Classification Can Use 0 and 1

Machine learning models often represent two classes using numerical labels such as 0 and 1.

The numbers are simply labels representing the categories.

Another problem could use the same idea differently:

Another Example — Fraud Detection

A bank may want to determine whether a transaction is fraudulent.

Because there are exactly two possible outcomes, this is another Binary Classification problem.

Binary vs Multi-Class Classification

The difference is simply the number of possible classes.

Simple Python Example

Here is a simple example using scikit-learn. The student result is represented using two labels: 0 and 1.

from sklearn.linear_model import LogisticRegression

X = [[1], [2], [3], [5], [6], [8]]
y = [0, 0, 0, 1, 1, 1]

model = LogisticRegression()

model.fit(X, y)

prediction = model.predict([[6]])

print(prediction)

In this example:

When the model predicts 1, we interpret that class as Pass.

Check Your Understanding