Algorithms for Decision Making

Livro digital

Título:
Algorithms for Decision Making

Autor:
Mykel J. Kochenderfer, Tim A. Wheeler, Kyle H. Wray

Categoria:
Tecnologia > IA

Doador:
Raffaello D. N.

Sinopse:
If you want more than a shallow tour of AI-flavored algorithms, this book signals its ambition immediately through a table of contents that moves from Representation, Inference, Parameter Learning, and Structure Learning into Simple Decisions, then expands into Markov decision processes, online planning, policy search, actor-critic methods, exploration and exploitation, belief-state planning, and multiagent systems. That progression makes the first impression concrete: the reader is being led from probabilistic foundations all the way to modern sequential decision-making machinery. The coverage is broad without being random. Early chapters frame uncertainty through Bayesian networks, sampling, and probabilistic models; the middle sections develop exact and approximate solution methods, value functions, Monte Carlo tree search, gradient-based optimization, and policy validation; later parts push into model uncertainty, imitation learning, particle filters, and planning under partial observability. The outline also reveals a practical differential: appendices on mathematical concepts, complexity, search algorithms, and neural representations help hold the heavier material together instead of leaving it as a stack of disconnected advanced topics. This is best suited to readers who already have some mathematical and programming maturity and want a serious map of decision making under uncertainty. Its strength is scope with structure: it ties reinforcement learning, probabilistic reasoning, planning, and multiagent thinking into one coherent framework that helps the reader understand not just individual algorithms, but the landscape they belong to.