Discretization

Discretization is the process of converting a continuous numeric variable into a set of discrete categories or bins - useful for simplifying data and for algorithms that work better with categorical inputs.

Why Discretize

# Converts:
Age = [22, 25, 31, 45, 52, 60]

# Into:
AgeGroup = ["Young", "Young", "Middle", "Middle", "Senior", "Senior"]

Common Discretization Methods

Equal-Width Binning:     Divides the range into bins of equal size
Equal-Frequency Binning: Divides data so each bin has roughly the same number of points
Custom Binning:          Uses domain knowledge to define meaningful bin edges

Implementing in Python

import pandas as pd

df = pd.DataFrame({
    "Age": [22, 25, 31, 45, 52, 60, 65, 70]
})

df["AgeGroup"] = pd.cut(
    df["Age"],
    bins=[0, 30, 50, 100],
    labels=["Young", "Middle", "Senior"]
)

print(df)

When to Use Discretization

Discretization can make patterns easier to interpret, reduce the effect of noise or outliers, and is sometimes required by algorithms that assume categorical inputs rather than continuous ones.

Coming Up Next

Next, you'll learn about Entropy, a key concept behind how Decision Trees decide where to split.