By Dimitri Bertsekas
The 3rd variation of the e-book is a completely rewritten model of the 1999 second version. New fabric was once integrated, a few of the previous fabric used to be discarded, and a wide section of the remaining used to be reorganized or revised.
This publication presents a accomplished and available presentation of algorithms for fixing non-stop optimization difficulties. It is determined by rigorous mathematical research, but additionally goals at an intuitive exposition that uses visualization the place attainable. It areas specific emphasis on glossy advancements, and their frequent functions in fields reminiscent of large-scale source allocation difficulties, sign processing, and computer learning.
The publication was once constructed via guideline at MIT, specializes in nonlinear and different sorts of optimization: iterative algorithms for restricted and unconstrained optimization, Lagrange multipliers and duality, huge scale difficulties, and the interface among non-stop and discrete optimization.
between its specific beneficial properties, the ebook: 1) presents large insurance of iterative optimization tools inside a unifying framework 2) presents an in depth therapy of inside element tools for linear programming three) covers intensive duality thought from either a variational and a geometrical/convex research standpoint four) comprises a lot new fabric on a couple of subject matters, akin to neural community education, large-scale optimization, sign processing, computing device studying, and optimum keep an eye on five) contains a huge variety of examples and workouts distinctive recommendations of a lot of that are published on the web. a lot supplementary/support fabric are available on the publisher's and the author's websites