دانلود کتاب هوش مصنوعی برای مهندسان – مبانی و پیاده سازی

This book presents basic knowledge and essential toolsets needed for people who want to step into artificial intelligence (AI). The book is especially suitable for those college students, graduate students, instructors, and IT hobbyists who have an engineering mindset. That is, it serves the idea of getting the job done quickly and neatly with an adequate understanding of why and how. It is designed to allow one to obtain big pictures for both AI and essential AI topics within the shortest amount of time. Based on the picture(s), suitable amounts of theoretical knowledge are contextualized to help the learner gain information about the most essential concepts and algorithms. These algorithms are introduced and formulated in a way that the learner can easily implement them for real-world applications with a small amount of effort. In short, you read it, you understand it, you try it, and you can solve it. The book is prepared by a learner for learners. Available learning materials like textbooks are diverse due to the broad scope and fast development of AI. There are inconsistent terminologies, distinct focuses of people from different application areas, inaccurate/misleading information from the Internet, and fast evolvement of available tools. As a result, a major problem of learners is not the lack of information but the opposite: there is too much information. A common frustration is the numerous gaps and conflicts that we can easily run into when we try to organize such knowledge to form a consistent and efficient literacy. We may need to read multiple books, take several actual or online courses, go through a lot of blogs or tutorials, and learn several tools via their user guides/manuals before we finally get what we want. Eventually, we may sadly find that much of the effort is unnecessary, while a lot of time is used to address the inconsistencies, gaps, and confusion arising during the above process. This book is prepared by a learner who has suffered from all the above issues for other learners who want to avoid such issues. This book, though titled Artificial Intelligence, is mostly devoted to numeric AI represented by machine learning algorithms, which predominate the so-called third wave/tide of AI. The most common and useful machine learning topics are selected and introduced. This includes introductions to topics that are needed to gain a basic understanding of machine learning, such as linear models, decision trees, Bayesian algorithms, and clustering algorithms, as well asmore advanced topics like deep learning and reinforcement learning. It attempts to cover the essential terms, basic/common algorithms, and useful tools that one may encounter in a typical journey for learning and performing contemporary AI. The book aims to strike a balance between being pragmatic and theoretical.Many AI learners, especially those in engineering applications, tend to solve a problem as quickly as possible, for example, using some AI code or libraries from the Internet. This usually works well considering the maturity of many AI tools. However, it may lead to unnecessary, inappropriate use of such tools and hinder further learning of the topics. On the contrary, some other learners try to start with intricate math and bottom-level computer science knowledge, which can easily discourage them and eventually turn out to be not needed in many cases. Each of the topics (or chapters) in this book adopts its own storyline, which may share a commonality with other topics while still maintaining uniqueness that stems from its own historical development and algorithmic nature. The book is designed for a typical undergraduate, graduate, or dual-listed course with a semester-based calendar. Accordingly, its 16 chapters present 16 essential topics for the 15 to 16 weeks in a typical university semester. The book has been tested in a dual-listed course containing both undergraduate and graduate students from engineering, in which each week had one lecture and one computer lab for each topic. Thus, additional teaching and learning materials, including lecture videos, notes, and code, are available. The book and associated materials have been proven to be able to assist a student with limited or even no knowledge in coding, data analytics, and machine learning to develop a basic and competing ability to understand, communicate, and implement AI. Due to the same reason, the book can also be considered by learners who are not an expert on or major in computer science but want to take 15 to 16 weeks or 45 cumulative hours (1.5 hours per session × 2 sessions per week × 15 weeks) to learn AI. This book can also be used as a reference book for more advanced AI practitioners who want to check the basic algorithms for reviewing a learned algorithm or learning new unfamiliar topics. This book starts with a chapter for the basics of AI, which both outlines a big AI picture by discussing why to look into AI, what is AI, the history of AI, AI vs. traditional engineering methods, as well as AI applications in engineering and other sectors, and presents the AI basics including the basic concepts, common algorithms, and challenges and issues. Next, in Chap. 2, common tools needed for implementing AI, including coding language and environment, data manipulation and visualization tools, machine learning and data analysis libraries, and deep learning packages, will be introduced. Following that, typical supervised learning topics, including linear models, decision trees, support vector machines, Bayesian algorithms, neural networks, and deep learning, will be presented in Chaps. 3–7. Ensemble learning will also be introduced in Chap. 9 based on them. After that, four typical unsupervised learning topics, i.e., clustering, dimensionality reduction, anomaly detection, and association rule learning, will be explored in Chaps. 10– 13. Reinforcement learning will be introduced using two chapters: Chap. 14 for the basics of reinforcement learning and value-based reinforcement learning, and Chap. 15 for policy-based reinforcement learning. Finally, Appendix A delivers appendices for more detailed information on AI’s foundational and ancillary knowledge in math, optimization (solvers), and evaluation metrics.