Introduction to Fraud and Anomaly Detection
This course is available only as a part of subscription plans.
Training duration: 4 hours (Hands-on)
Develop good features (recency, frequency, and monetary value as well as categorical transformations) for detecting and preventing fraud
Identify anomalies using statistical techniques like z-scores, robust z-scores, Mahalanobis distances, k-nearest neighbors (k-NN), and local outlier factor (LOF)
Identify anomalies using machines learning approaches like isolation forests and classifier adjusted density estimation (CADE)
Visualize these anomalies identified by the above approaches
Aric LaBarr, PhD
Aric LaBarr, PhD
Module 1: Introduction
Lesson 1.1 - Who am I
Lesson 1.2 - What are Anomalies
Lesson 1.3 - Anomaly Detection Analytical Framework
Module 2: Data Preparation
Lesson 2.1 - Feature Engineering
Lesson 2.2 - Recency and Frequency
Lesson 2.3 - Periodic Means
Lesson 2.4 - Categorical Feature Engineering
Module 3: Probability and Statistical Approaches
Lesson 3.1 - Benford's Law
Lesson 3.2 - Z-scores and Robust Z-scores
Lesson 3.3 - IQR Rule and Its Adjustment
Lesson 3.4 - Mahalanobis Distances and Robust Mahalanobis
Module 4: Machine Learning Approaches
Lesson 4.1 - k-Nearest Neighbors (k-NN)
Lesson 4.2 - Local Outlier Factor (LOF)
Lesson 4.3 - Isolation Forests
Lesson 4.4 - Classifier-Adjusted Density Estimation (CADE)
Lesson 4.5 - One-Class Support Vector Machines (SVM)
Introductory knowledge to statistics to understand means and standard deviations
Introduction to basic machine learning to grasp the concepts of the advanced anomaly detection