Feature Selection Techniques

Feature selection is the process of choosing the most relevant columns from your dataset so the model focuses only on data that improves predictions and ignores the rest.

Why feature selection matters

• Removing irrelevant features reduces noise and overfitting.
• Fewer features mean faster training and simpler models.
• It improves model accuracy by removing misleading information.

Correlation-based selection

Highly correlated features often carry duplicate information, so you can remove one.

correlation = df.corr()
print(correlation)

Checking correlation values helps you spot features that are almost saying the same thing, so you can drop the less useful one.

Selecting features using a model

Some models can rank features by importance, helping you decide which ones matter most.

from sklearn.ensemble import RandomForestClassifier

model = RandomForestClassifier()
model.fit(X, y)

print(model.feature_importances_)

feature_importances_ shows how much each feature contributed to the model's predictions, so you can keep the strong ones and drop the weak ones.

Good feature selection leads to a model that is simpler, faster, and more accurate, instead of one overloaded with unnecessary data.