This book presents an introduction to the calibration (estimation of parameters) of the Schwartz (1997) reduced-forrn, no-arbirrage two factor model by applying a combination of the Kalman filter and the maximum log-likelihood method knows as the predictive error decomposition. This book is written in such a way that a reader with primary tools in stochastic calculus and optimization (mainly the maximum log-Iikelihood method) can find the necessary tools for doing its reading without problems and understand the essential elements of the methodology.