Feature Engineering Techniques

Here are some of the most widely used Feature Engineering techniques, along with simple Python examples for each.

Encoding Categorical Variables

import pandas as pd

df = pd.DataFrame({"City": ["Noida", "Delhi", "Noida"]})

# One-hot encoding
df_encoded = pd.get_dummies(df, columns=["City"])

Scaling Numeric Features

from sklearn.preprocessing import MinMaxScaler

scaler = MinMaxScaler()
df["ScaledIncome"] = scaler.fit_transform(df[["Income"]])

Binning Continuous Values

df["AgeGroup"] = pd.cut(df["Age"], bins=[0, 18, 35, 60, 100],
                        labels=["Teen", "Young Adult", "Adult", "Senior"])

Extracting Date Parts

df["OrderDate"] = pd.to_datetime(df["OrderDate"])
df["Month"] = df["OrderDate"].dt.month
df["Weekday"] = df["OrderDate"].dt.day_name()

Creating Ratio and Interaction Features

df["ExpensePerPerson"] = df["TotalExpense"] / df["FamilySize"]
df["PriceTimesQuantity"] = df["Price"] * df["Quantity"]

There's no universal "best" technique - the right choice depends on the model you're using and the nature of the data, so it's common to try several techniques and compare their impact on model performance.

Coming Up Next

Next, you'll shift focus to the basics of Probability - a foundation that underpins much of statistics and machine learning.