Build a Model for Anomaly Detection in Time Series Data
This course will teach you techniques to build a model for anomaly detection on your own time series dataset.
What you'll learn
In the real world, time series data is one of the most used and researched types of data, and anomaly detection in it has innumerable uses ranging from detecting fraud transactions, uncovering fraudulent insurance claims, and even detecting critical equipment failures.
In this course, Build a Model for Anomaly Detection in Time Series Data, you'll learn different techniques to build a model for anomaly detection on your very own time series dataset.
First, you’ll be introduced to time series data and its different components, what anomaly detection means when it pertains to a time series dataset, and its importance.
Next, you’ll discover different techniques with which to build models that detect anomalies in time series datasets.
Finally, you’ll learn how to deal with the anomalies that you previously detected in your dataset.
When you’re finished with this course, you’ll have the skills and knowledge needed to explore, clean, prepare, and detect anomalies on your own time series dataset.
Table of contents
- Module Overview 1m
- STL Decomposition 2m
- Classification and Regression Trees (CART) 4m
- Clustering-based Anomaly Detection 3m
- Anomaly Detection Using Autoencoders 3m
- Demo: Introduction to the Problem and Dataset 2m
- Demo: Exploratory Data Analysis and Data Cleaning 8m
- Demo: Data Preprocessing and Dimensionality Reduction 6m
- Demo: Building a Model for Anomaly Detection 6m
- Module Summary 1m