Overview

Google, IBM & Meta Certificates – 40% Off

One Coursera Plus subscription covers most Professional Certificates on Coursera.

This course teaches you how to design and implement a complete, production-ready anomaly detection system for both tabular and time-series data. You will start by distinguishing point, contextual, and collective anomalies and understanding why fixed thresholds often fail in noisy, evolving real-world environments. From there, you will build strong statistical baselines and engineer features that make abnormal patterns more detectable across different data modalities. You will then train and interpret a range of unsupervised machine learning methods, including Isolation Forest, Local Outlier Factor, One-Class SVM, KNN-based outlier scoring, and autoencoder-based models for high-dimensional and seasonal time-series data. You will learn how to rigorously evaluate detectors under class imbalance using precision, recall, PR-AUC, and top-K precision, and how to compare ML approaches against statistical baselines. The course guides you through constructing an ensemble pipeline that normalizes and combines outputs from multiple detectors into a unified anomaly score with confidence, and through adding explainability via feature attributions, per-signal breakdowns, visualizations, and human-readable alert summaries. Finally, you will deploy your anomaly detection pipeline as a FastAPI service with rich scoring outputs, detector breakdowns, and integration points for downstream systems. You will design monitoring and continuous improvement workflows to handle concept drift, noisy and evolving data, analyst feedback loops, and controlled retraining and model updates in production, ensuring your anomaly detection system remains robust and valuable over time. Disclaimer: This is an independent educational resource created by Board Infinity for informational and educational purposes only. This course is not affiliated with, endorsed by, sponsored by, or officially associated with any company, organization, or certification body unless explicitly stated. The content provided is based on industry knowledge and best practices but does not constitute official training material for any specific employer or certification program. All company names, trademarks, service marks, and logos referenced are the property of their respective owners and are used solely for educational identification and comparison purposes.

Syllabus

Foundations & Building the Anomaly Detector
- This module builds the foundation for a practical anomaly detection system by showing how to move beyond simple threshold rules and toward more capable statistical and machine learning approaches. You will examine how different types of anomalies appear in fraud, security, infrastructure, and sensor data, engineer features that make unusual behavior easier to detect, and train unsupervised detectors on real data. By the end of the module, you will have a strong base for selecting and comparing detection methods across tabular and time-series settings.
Multi-Detector Pipeline, Evaluation & Deployment
- This module helps you turn separate anomaly detection methods into a production-oriented system that can score incoming data, explain alerts, and support operational decision-making. You will combine detector outputs into a unified anomaly score, evaluate performance under severe class imbalance, deploy the pipeline as a FastAPI service, and plan for monitoring, feedback, and drift over time. These skills mirror the real challenges faced in fraud detection, observability, security, and other high-stakes monitoring environments.