Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

CodeSignal

Evaluation Metrics & Advanced Techniques for Imbalanced Data

via CodeSignal

Go to class Write review

Overview

This course focuses on evaluating models with imbalanced data, and explores advanced techniques including cost-sensitive learning, ensemble methods, and anomaly detection for extreme imbalance. You'll learn appropriate evaluation strategies specifically designed for imbalanced datasets, helping you choose the right metrics and avoid common pitfalls.

Syllabus

  • Unit 1: Comprehensive Evaluation Strategies for Imbalanced Data
    • Beyond Accuracy Metrics for Imbalanced Data
    • Extracting Insights from Confusion Matrices
    • Fixing Parameter Order in Evaluation Metrics
    • Generating Classification Reports for Imbalanced Data
  • Unit 2: Balanced Logistic Regression: Improving Minority Class Detection with Class Weights
    • Balancing the Scales with Class Weights
    • Fixing Parameter Names for Class Weights
    • Comparing Model Performance with Classification Reports
    • Building a Complete Class Weights Comparison
    • Custom Class Weights Under the Hood
  • Unit 3: Ensemble Methods for Imbalanced Data
    • Implementing BalancedBaggingClassifier for Imbalanced Data
    • Fixing Imports for Ensemble Methods
    • Fixing Prediction Sequence in Ensemble Models
  • Unit 4: Anomaly Detection for Extremely Imbalanced Data
    • Training Anomaly Detectors with Majority Data
    • Fixing Isolation Forest Prediction Conversion
    • Setting the Right Contamination Parameter
    • Evaluating Anomaly Detection Model Performance
    • Training Isolation Forest on Majority Class
    • Building a Complete Anomaly Detection Pipeline

Reviews

Start your review of Evaluation Metrics & Advanced Techniques for Imbalanced Data

Go to class

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Sign up for free
Someone learning on their laptop while sitting on the floor.