Chapter 1: What is Anomaly Detection?
Chapter Goal: Introduce reader to the task of anomaly detection, where it's used, why it's important, as well as the different types of “anomaly detection” there are.
No of pages: 30
Sub -Topics
1. What is an anomaly
2. Use cases today
3. Different types of anomalies
Chapter 2: Traditional Methods of Anomaly Detection
Chapter Goal: Introduce reader to a couple high performing traditional methods of anomaly detection in Scikit-Learn. Evaluation metrics are performed on both (will be set as the benchmark of comparison for the deep learning models later on)
No of pages: 50
Sub - Topics
1. Isolation Forest
2. One class support vector machine
3. Mahalanobis distance based anomaly detection
Chapter 3: Intro to Keras and PyTorch
Chapter Goal: Introduce reader to deep learning and how to build, train a basic model in both Keras and in PyTorch. Additionally, perform evaluation metrics on both. Also, discuss the various deep learning models that can be applied to semi-supervised and unsupervised anomaly detection.
No of pages : 40
Sub - Topics:
1.What is deep learning?
2. Intro to Keras: simple classifier model
3. Intro to PyTorch: simple classifier model
4. How can we apply deep learning to anomaly detection?
Chapter 4: Autoencoders
Chapter Goal: Introduce reader to several autoencoders and how they can perform anomaly detection in both semi-supervised and unsupervised anomaly detection.
No of pages: 40
Sub - Topics:
1. What are autoencoders?
2. Basic autoencoder
3. Denoising autoencoder
4. Variational autoencoder
5. Summary of autoencoders as a model
Chapter 5: Boltzmann Machines
Chapter Goal: Introduce reader to a restricted Boltzmann machine, deep Boltzmann machine, and a deep belief network.
No of pages: 30
Sub - Topics:
1. What is a Boltzmann machine?
2. RBM
3. DBM
4. DBN
5. Summary of the models
Chapter 6: Time-Series Anomaly Detection
Chapter Goal: Introduce reader to RNNs and LSTMs for time series anomaly detection.
No of pages:
Sub - Topics: 30
1. What is a time series and how do we detect anomalies in that?
2. What is an RNN
3. RNN application
4. What is an LSTM?
5. LSTM application
6. Summary of the models
Chapter 7: Temporal Convolutional Network
Chapter Goal: Introduce reader to the TCN and how it can be used in anomaly detection.
No of pages: 30
Sub - Topics:
1. What is a TCN?
2. Encoder-Decoder TCN
3. Dilated TCN
4. Summary of models
Chapter 8: Practical Use Cases of Anomaly Detection
Chapter Goal: Illustrate common use cases.
No of pages: 30
Sub - Topics:
1. Use cases
Appendix A: Introduction to Keras
Chapter Goal: Introduce reader to the Keras
No of pages: 30
Sub - Topics:
1. What is a Keras?
2. How to use it
Appendix B: Introduction to PyTorch
Chapter Goal: Introduce reader to the PyTorch
No of pages: 30
Sub - Topics:
1. What is a PyTorch?
2. How to use it