Pricing anomalies are when a stock is priced differently than how a model predicts it will be priced. Being able to accurately predict such anomalies aids regulators to detect stock market manipulation and prevent against fraud. The aim of this project is to design a stock pricing anomaly detector using Azure technologies.
Detecting abnormal behaviour has long been a useful tactic to prevent against crimes such as fraud, manipulation, system performance, monitoring, etc. One such case is detection of stock market manipulation. Machine learning is an effective method to predict anomalous points in time series data. Azure Anomaly Detector can figure out abnormal points from time series data without any machine learning coding based on its pre-trained machine learning model. Using Azure Anomaly Detector, a stock price anomaly detection solution is designed. The tool uses time series data of the pricing of a stock to predict anomalies in data and prevent market manipulation.
Primary Azure Technologies: Anomaly Detector, Azure Cognitive Services
Industry: Fin-Tech