01 / Reading Guide
Reading Guide
Who this book is for
For advanced undergraduate students who want to turn an optimization model into a working algorithm and understand the resulting trade-offs. It is useful across mathematics, engineering and computer science. The book offers a computational complement to a convex-optimization text, with broader attention to algorithms used when the objective or feasible region is difficult.
Prerequisites
Multivariable calculus, linear algebra, elementary probability and programming. Be able to compute a gradient and interpret an objective function before comparing numerical optimization methods.
What it covers
Local search, derivative-based and derivative-free methods, constraints, surrogate models and uncertainty are connected to implementation. Figures and code help compare the assumptions behind different methods. The treatment encourages matching the algorithm to the information available about a problem rather than selecting a method by name alone.
How to use it
Start with a small problem whose behaviour you can plot. Implement a simple baseline, then change one algorithmic choice at a time. Record the initial point, stopping rule, evaluation count and achieved objective so comparisons are reproducible. Check feasibility separately from improvement in the objective, and use several starts when a local solution is all the method can guarantee.
Edition and sources
This record identifies the 2019 first edition, hardcover ISBN 9780262039420, confirmed by MIT Press. The author site supplies a separate first-edition PDF. Its 2027 second-edition preview is a different version and is not the edition described here.
