Course

Course Summary
Credit Type:
Course
ACE ID:
STAT-0033
Organization:
Statistics.com
Location:
Classroom-based
Length:
4 weeks (60 hours)
Dates Offered:
Credit Recommendation & Competencies
Level Credits (SH) Subject
Upper-Division Baccalaureate 3 Data Mining, Statistics, or Computer Science
Description

Objective:

The course objective is to learn how to examine data with the goal of detecting anomalies or abnormal instances, a task critical in a wide range of applications ranging from fraud detection to surveillance. Students will understand the different aspects that affect how a problem can be formulated, the techniques applicable for each formulation, and knowledge of some real-world applications in which they are most effective.

Learning Outcomes:

  • Determine how to apply a supervised learning algorithm to a classification problem for anomaly detection
  • Explain the limitations of supervised learning for anomaly detection
  • Explain the advantages and disadvantages of various statistical methods for identifying anomalies in the absence of labels
  • Apply and assess a nearest-neighbor algorithm for identifying anomalies in the absence of labels
  • Practice applying the various techniques to different problems in different domains
  • Use Python to implement such a model
  • Use characteristics of the data and its originating domain to make judgments about which methods work best to identify anomalies

General Topics:

  • Different aspects of anomalies
  • Classification-based approaches
  • Clustering and nearest neighbor
  • Information-theoretic methods
  • Spectral techniques
  • Credit-card fraud, intrusion detection and surveillance
Instruction & Assessment

Instructional Strategies:

  • Case Studies
  • Classroom Exercise
  • Computer Based Training
  • Discussion
  • Lectures
  • Practical Exercises

Methods of Assessment:

  • Case Studies
  • Quizzes

Minimum Passing Score:

70%
Supplemental Materials