This book focuses on the advancements of Estimation of Distribution Algorithms (EDAs) that perform optimization via building and sampling probabilistic models of promising solutions. Initial chapters contain brief introduction to investigated areas genetic algorithms, probabilistic models, and optimization via probabilistic models. Different disadvantages of classical genetic algorithms are highlighted and the utilization of probabilistic models in evolutionary computation is justified. Main part of the book is devoted to the development of advanced EDAs for application areas where present EDAs are unapplicable or ineffective. Multiple efficiency enhancement techniques are discussed. An advanced tree-based probabilistic model is developed to allow for solving optimization problems with mixed continuous-discrete variables. Coarse-grained and fine-grained parallel EDAs are implemented for time-critical applications. Utilization of prior knowledge about the problem is proposed andempirically investigated. And, the concept of Pareto fronts is employed to design multiobjective EDAs.