Knowledge about artificial intelligence should be available to everyone. This will create equal opportunities and encourage people from different backgrounds to take part in advancing AI. Through the digital transformation of our societies, we will change the future of learning, healthcare, public administration. The digitalization of the labor market will have a positive impact on the economy and drive a change towards more successful and productive industries.
In this article, we share with you five books that are available to download for free. We hope that they will motivate you to improve your knowledge and implement some of the techniques on the real-word problems that you may encounter.
Machine Learning, Neural and Statistical Classification
This book was written by D. Michie, D.J. Spiegelhalter, C.C. Taylor in 1994. Even though it was written a long time ago, it covers all the important classification topics and algorithms that we use today. The authors explore classification techniques while assessing their usability, advantages, and disadvantages for different usage cases. Throughout this book, the authors used challenging data-sets and considered the applicability of different classification algorithms on realistic industrial problems. While the approach of this book is statistical, the emphasis is on concepts and their applications.
The Elements of Statistical Learning
This book was written by Trevor Hastie, Robert Tibshirani, Jerome Friedman, who are professors of statistics at Stanford University. This book is widely known as the Bible of Artificial Intelligence. If you are interested in applications of data mining in science and industry, this is the right book for you. It doesn’t emphasize mathematical concepts, rather it focuses on the applications of algorithms in the industry. They give many examples of every technique that they explain. This book covers a broad range of topics, from supervised learning to unsupervised learning. Implementation of algorithms is done in R language and the code is available on their official website.
Ethical Artificial Intelligence
Ethical Artificial Intelligence was written by Bill Hibbard. It is aimed at people who are interested in a scientific approach towards answering the question of how our future might be affected by AI. It tackles the problem of building ethical AI systems, such as whether the AI systems may disrupt the initial purposes of their development as both they and humanity evolve. Apart from pointing out the problems and discussing them, this book also makes the case for various approaches that can be used to solve the mentioned problems. One of the interesting topics that this book covers is the possibility of inspiring anger and hatred among humans, as a result of unethically designed AI systems. It argues that AI would defeat humans not with its brawn, but with brains. In this scenario, humans might not be able to realize that AI is the cause of their arguments.
Learning From Data
Learning From Data is used as teaching material on many universities especially for short introductory courses on Statistical Machine Learning. Strengths and advantages of different algorithms are pointed out, which is very beneficial for all learners who want to learn the best uses cases for particular techniques. On their official website, you can find additional materials such as video lectures that cover different parts of the book and lecture slides. On this website, you can also find a Handwritten Digits dataset that is extremely useful for experimenting with some of the algorithms discussed in the book. You can download the raw data set as well as the data that already have extracted features that you can use for training. Throughout the book, you will find well-designed exercises that will enhance your understanding of a particular topic. This book has a good balance between the theoretical and the practical and it is not mathematics heavy. The authors of Learning From Data book are Yaser S. Abu-Mostafa, Malik Magdon-Ismail, and Hsuan-Tien Lin.
Deep Learning
The Deep Learning textbook is a beginner-friendly resource for anyone interested in entering the field of machine learning in general and deep learning in particular. It covers a wide range of topics, including very advanced ones, but it also has fundamental mathematical concepts explained in the first couple of chapters. Relevant concepts in linear algebra, probability theory and information theory, numerical computation, and machine learning are explained in details and they serve as a good foundation for following the rest of the book easily, which is especially important for beginners in this area. The authors are Ian Goodfellow, Yoshua Bengio and Aaron Courville, who are research leaders in the field of deep learning. Basically, if you try reading any of the important papers on deep learning, you’ll realize that most of them were written by these authors.
The online version of the book is available for free on their official website. The website also provides a variety of supplementary materials, such as exercises and lecture slides, that should be useful to both readers and instructors.
