🔢 02 — Numeric Features

Numeric features often need scaling, transformation, or binning.

---

Scaling

Use scaling for distance-based and linear models.

from sklearn.preprocessing import StandardScaler, MinMaxScaler

StandardScaler()
MinMaxScaler()

Models that often benefit:

• Logistic Regression
• Linear Regression with regularization
• KNN
• SVM
• Neural Networks

Tree models usually do not require scaling.

---

Log Transform

Useful for right-skewed values like income or spend.

import numpy as np

df["log_spend"] = np.log1p(df["spend"])

log1p handles zero safely.

---

Binning

df["age_group"] = pd.cut(
    df["age"],
    bins=[0, 25, 40, 60, 100],
    labels=["young", "adult", "middle", "senior"]
)

Binning can improve interpretability but may lose detail.

---

Interaction Features

df["revenue_per_order"] = df["revenue"] / df["order_count"]
df["price_per_unit"] = df["revenue"] / df["quantity"]

Create interactions when domain logic supports them.

---

Next

➡️ 03-categorical-features