🕵️ Analytics Detective Academy

Master Anomaly Detection & Market Basket Analysis

⭐ Points
0
🏆 Level
1
🔥 Streak
0
🎯 Accuracy
0%
🔍 Anomaly Detection: Credit Card Fraud
You're monitoring transactions at a bank. Most are normal, but some are fraudulent. Can you spot the anomalies? Look for unusual amounts, times, or patterns!
💡 Key Concept: One-Class Classification
We only train on "normal" data, then flag anything that doesn't fit the pattern. Look for transactions that deviate significantly from the typical behavior!
💡 Hint: Check for transactions with unusually high amounts or occurring at odd times!
🛒 Market Basket Analysis: Retail Store
Analyze customer purchases to find associations. Which products are frequently bought together? Build rules with high confidence and lift!
Available Products
📊 Metrics to Remember
Support: How often items appear together
Confidence: P(B|A) - probability of B given A
Lift: How much more likely B is purchased with A
Create Association Rule

Select products to create a rule...

Support -
Confidence -
Lift -
💡 Hint: Look for classic combinations like bread & butter, or laptop & mouse!
🎯 Master Detective Challenge
An e-commerce site is experiencing fraud. Use both anomaly detection AND market basket analysis to identify suspicious purchasing patterns!
🔗 Combining Techniques
Real-world problems often need multiple techniques. Here, use associations to understand normal buying patterns, then detect anomalies that violate these patterns!