This book covers analytic methods to solve one-dimensional and multi-dimensional problems with or without possible constraints, iterative numerical techniques based on the gradient calculation or its estimation, and n...
详细信息
ISBN:
(数字)9783031863264
ISBN:
(纸本)9783031863257;9783031863288
This book covers analytic methods to solve one-dimensional and multi-dimensional problems with or without possible constraints, iterative numerical techniques based on the gradient calculation or its estimation, and numerical methods that do not require the knowledge of gradient and use only comparative iterative tests. This book provides the reader with a basic introduction to some traditional parameter optimization techniques. The presented problems and their solution methods represent a core of the parameter optimization reign since the 17th century to the 1970s. Linear and integer programming via the simplex table is also introduced. Two simple selected problems that are solved using dynamic programming principles are also given to the reader. A general approach to constraints via penalty and barrier functions is introduced. A concise introduction to the decision and game theory concludes the book.
The book does not intend to provide the reader with a rigorous mathematic derivation of the presented methods. Its aim is instead to bring to the attention essential optimization tools for practitioners and undergraduate students and introduce selected well-established techniques to them when optimizing parameters of various models. Each method is described theoretically and supported by one or more numerical examples that vary from academic ones, through business economics to funny real-world problems that attract a broad audience. A sketch of Matlab code also follows numerical-based techniques. The author believes that the book finds its place in the libraries of many undergraduate students of various technical study programs and modern, thoughtful people worldwide, regardless of their expertise.
This book presents techniques to render photo-realistic images by programming the Graphics Processing Unit (GPU). We discuss effects such as mirror reflections, refractions, caustics, diffuse or glossy indirect illumi...
详细信息
ISBN:
(数字)9783031795251
ISBN:
(纸本)9783031795244
This book presents techniques to render photo-realistic images by programming the Graphics Processing Unit (GPU). We discuss effects such as mirror reflections, refractions, caustics, diffuse or glossy indirect illumination, radiosity, single or multiple scattering in participating media, tone reproduction, glow, and depth of field. The book targets game developers, graphics programmers, and also students with some basic understanding of computer graphics algorithms, rendering APIs like Direct3D or OpenGL, and shader programming. In order to make the book self-contained, the most important concepts of local illumination and global illumination rendering, graphics hardware, and Direct3D/HLSL programming are reviewed in the first chapters. After these introductory chapters we warm up with simple methods including shadow and environment mapping, then we move on toward advanced concepts aiming at global illumination rendering. Since it would have been impossible to give a rigorous review of all approaches proposed in this field, we go into the details of just a few methods solving each particular global illumination effect. However, a short discussion of the state of the art and links to the bibliography are also provided to refer the interested reader to techniques that are not detailed in this book. The implementation of the selected methods is also presented in HLSL, and we discuss their observed performance, merits, and disadvantages. In the last chapter, we also review how these techniques can be integrated in an advanced game engine and present case studies of their exploitation in games. Having gone through this book, the reader will have an overview of the state of the art, will be able to apply and improve these techniques, and most importantly, will be capable of developing brand new GPU algorithms. Table of Contents: Global Illumintation Rendering / Local Illumination Rendering Pipeline of GPUs / Programming and controlling GPUs / Simple Improvements of the L
This book constitutes the refereed proceedings of the 7th International Conference on Evolutionary Multi-Criterion Optimization, EMO 2013 held in Sheffield, UK, in March 2013. The 57 revised full papers presented were...
详细信息
ISBN:
(数字)9783642371400
ISBN:
(纸本)9783642371394
This book constitutes the refereed proceedings of the 7th International Conference on Evolutionary Multi-Criterion Optimization, EMO 2013 held in Sheffield, UK, in March 2013. The 57 revised full papers presented were carefully reviewed and selected from 98 submissions. The papers are grouped in topical sections on plenary talks; new horizons; indicator-based methods; aspects of algorithm design; pareto-based methods; hybrid MCDA; decomposition-based methods; classical MCDA; exploratory problem analysis; product and process applications; aerospace and automotive applications; further real-world applications; and under-explored challenges.
This book constitutes the proceedings of the 8th International Conference on Swarm Intelligence, held in Brussels, Belgium, in September 2012.This volume contains 15 full papers, 20 short papers, and 7 extended abstra...
详细信息
ISBN:
(数字)9783642326509
ISBN:
(纸本)9783642326493
This book constitutes the proceedings of the 8th International Conference on Swarm Intelligence, held in Brussels, Belgium, in September 2012.
This volume contains 15 full papers, 20 short papers, and 7 extended abstracts carefully selected out of 81 submissions. The papers cover various topics of swarm intelligence.
To promote the responsible development and use of data-driven technologies –such as machine learning and artificial intelligence– principles of trustworthiness, accountability and fairness should be followed. The qu...
详细信息
To promote the responsible development and use of data-driven technologies –such as machine learning and artificial intelligence– principles of trustworthiness, accountability and fairness should be followed. The quality of the dataset on which these applications rely, is crucial to achieve compliance with the required ethical principles. Quantitative approaches to measure data quality are abundant in the literature and among practitioners, however they are not sufficient to cover all the principles and ethical challenges *** this paper, we show that complementing data quality with measurable dimensions of data documentation and of data balance helps to cover a wider range of ethical challenges connected to the use of datasets in algorithms. A synthetic report of the metrics applied (the Extended Data Brief) and a set of Risk Labels for the Ethical Challenges provide a practical overview of the potential ethical harms due to data composition. We believe that the proposed data labelling scheme will enable practitioners to improve the overall quality of datasets and to build more responsible data-driven software systems.
暂无评论