Machine Learning for High-Risk Applications

The book is a comprehensive guide on the application of artificial intelligence and machine learning (AI/ML) aimed at reducing risks for modern businesses associated with the use of these technologies. It covers the fundamentals of risk management and cybersecurity, regulations, accountability for product quality based on ML, as well as explainable models and methods for their verification, including the new NIST AI risk management framework. The reader is offered an in-depth look at programming using Python with detailed examples for structured and unstructured data. Special attention is paid to explainable boosting machines, the XGBoost library, and methods for improving the quality of ML models. Practical advice based on real experience is presented on how to successfully work with high-risk applications. Practical examples illustrate the importance and complexity of implementing ML in various industries.