Data series anomaly detection is a crucial and challenging problem in data science . The goal is to train an algorithm to automatically detect instances (such as data series) as an anomaly. One class learning is one way to tackle the anomaly detection task. It requires to have a dataset of normal data series and train a model to recognize them. If abnormal labels are provided, classification algorithms can be used to solve an anomaly detection task. In this project, you will tackle both tasks.