一、課程說明 (Course Description)

This course aims to provide an introduction of various machine learning
techniques to detect anomalies in data for several different kinds of
applications. The course will cover three types of anomaly detection problems,
time-series signal, image, and video anomaly detections. We will start with
introducing traditional machine learning techniques for anomaly detection, and
the latter part of the course will focus on deep learning approaches to anomaly
detection. We will discuss the recent advances of deep learning techniques for
anomaly detection. This course will focus on the application of anomaly detection
in smart manufacturing, video surveillance, and cybersecurity.

二、指定用書 (Text Books)

1. Anomaly Detection Principles and Algorithms, Kishan G. Mehrotra,
Chilukuri K. Mohan, and HuaMing Huang, Springer Publisher, 2017
2. some recent papers assigned for reading
3. Lecture slides

三、教學進度 (Syllabus)

1. Introduction to Anomaly Detection
2. Distance-based Anomaly Detection
3. Clustering-based Anomaly Detection
4. Model-based Anomaly Detection
5. Anomaly Detection for Time-Series Data
6. Introduction to Deep Learning
7. Practical DNN training techniques
8. Generative Models
9. Deep Learning for Image Anomaly Detection
10. Deep Learning for Video Anomaly Detection
11. Deep Learning for Time-Series Anomaly Detection
12. Zero-shot and Few-shot Anomaly Detection
13. Guest lectures
14. Final Project Presentation

四、教學方式 (Teaching Method)

Lectures (instructors and guest speakers) and interactive discussion

五、參考書籍 (References)

Beginning Anomaly Detection Using Python-based Deep Learning: With Keras and
PyTorch, Sridhar Alla and Suman Kalyan Adari, 2019

六、成績考核 (Evaluation)

Homeworks 30%
Midterm Exam 30%
Term project (team-based) 30%
Class Attendance 5%
Class Participation 5%

七、可連結之網頁位址:
eeclass: https://eeclass.nthu.edu.tw/