An Introduction to Sequential Monte Carlo

More about the book

This book offers a comprehensive introduction to Sequential Monte Carlo (SMC) methods, commonly known as particle filters, which are essential for sequential data analysis across various fields, including signal processing, epidemiology, machine learning, and robotics. It covers the theoretical foundations, computational implementation, and methodologies, framing SMC algorithms within a general framework that incorporates concepts like Feynman-Kac distributions and techniques such as importance sampling and resampling. The text emphasizes sequential learning of state-space models, a key application of SMC methods, while also addressing recent advancements in parameter estimation and simulation of complex probability distributions. Designed as both a graduate textbook and a reference work, each chapter includes exercises, a comprehensive bibliography, and a "Python corner" for practical implementation. Additionally, the book is accompanied by an open-source Python library that implements all discussed algorithms and provides the programs used for numerical experiments. The structured content spans various topics, including state-space models, Markov processes, particle filtering, and advanced concepts in SMC, making it a valuable resource for both students and practitioners in the field.