Gilbert Strang Knihy

Offering a fresh perspective on linear algebra, this book caters to learners of all levels, guiding them from fundamental concepts to advanced topics like singular value decomposition. Its accessible approach ensures that readers can grasp complex ideas easily, making it a valuable resource for students and enthusiasts alike.

Linear algebra is something all mathematics undergraduates and many other students, in subjects ranging from engineering to economics, have to learn. The fifth edition of this hugely successful textbook retains all the qualities of earlier editions, while at the same time seeing numerous minor improvements and major additions. The latter include: • A new chapter on singular values and singular vectors, including ways to analyze a matrix of data • A revised chapter on computing in linear algebra, with professional-level algorithms and code that can be downloaded for a variety of languages • A new section on linear algebra and cryptography • A new chapter on linear algebra in probability and statistics. A dedicated and active website also offers solutions to exercises as well as new exercises from many different sources (including practice problems, exams, and development of textbook examples), plus codes in MATLAB®, Julia, and Python.

Focusing on the intersection of linear algebra and deep learning, this textbook by Professor Gilbert Strang offers a comprehensive course that integrates essential mathematical concepts with practical applications in neural networks. It covers key topics such as the four fundamental subspaces, singular value decompositions, and optimization techniques, along with foundational elements of probability and statistics. The text is designed to be both accessible and rigorous, making it an invaluable resource for students eager to understand how linear algebra underpins modern data learning techniques.

Eine wesentliche Notwendigkeit für heutige Studenten und Leser besteht darin, von den herkömmlichen formelbasierten Kursen abzukommen und zu rechnergestützten Kursen überzugehen. Das Ziel dieses jetzt auch endlich in deutscher Version erhältlichen Buches ist es, sowohl angewandte Mathematik als auch Ingenieurmathematik so darzustellen, wie sie heutzutage tatsächlich Anwendung finden! Dieses Buch entstand aus dem Kurs zu wissenschaftlichem Rechnen, der seit 20 Jahren am Massachusetts Institute of Technology abgehalten wird. Das Buch versucht, Konzepte und Algorithmen für den Leser zusammenzuführen. Die Autoren beginnen mit der angewandten linearen Algebra, einem bei vielen Lesern zu kurz gekommenen Gebiet, welches aber ein wesentliches Werkzeug für das wissenschaftliche Rechnen und seine Anwendungen ist. Anschließend entwickeln sie die Methoden der finiten Differenzen und finiten Elemente, stets mit Hinblick auf die angewandte Mathematik, um dieses Gebiet mit Anwendungen in zahlreichen Wissensgebieten in Verbindung zu bringen. Studenten, Dozenten und Forscher werden dieses Buch gleichermaßen mit großem Gewinn lesen.

Gilbert Strang's clear, direct style and detailed, intensive explanations make this textbook ideal as both a course companion and for self-study. Single variable and multivariable calculus are covered in depth. Key examples of the application of calculus to areas such as physics, engineering and economics are included in order to enhance students' understanding. New to the third edition is a chapter on the 'Highlights of calculus', which accompanies the popular video lectures by the author on MIT's OpenCourseWare. These can be accessed from math.mit.edu/~gs.

Covering the entire spectrum of computational science and engineering, this book delves into both modeling techniques and solution methods, including analytical and numerical approaches. It serves as a comprehensive resource for understanding the intricacies of computational methods and their applications in various scientific and engineering fields.