🚨 Anomaly Detection

Finding the needle in the haystack.

🐍 Python Implementation

We use IsolationForest to isolate weird points (like credit card fraud).

from sklearn.ensemble import IsolationForest

# Mostly normal data near 0
X = [[-1.1], [0.3], [0.5], [100]] # 100 is highly anomalous!

# Create the alarm system
iso = IsolationForest(random_state=42)
predictions = iso.fit_predict(X)

# 1 = Normal, -1 = Anomaly
print("Anomaly predictions:", predictions)