Talk:Practical Guide to Gaussian Processes

On this new book
This book is a translation of the German Wikipedia article of Gaussian Process (see Gaußprozess. I am the main author of this German article. Since I am no english native speaker I used for the translation the tools deepl.com and in some cases chatGPT. I hope the text reads smoothly. Physikinger (discuss • contribs) 22:27, 21 March 2023 (UTC)


 * Hi @Physikinger! Unfortunately, encyclopedic articles like this do not fall under the Wikibooks scope. Unless it can be turned into a book, it will likely need to be deleted. For more information on what Wikibooks allows, please see What is Wikibooks. Thanks! —Kittycataclysm (discuss • contribs) 00:25, 22 March 2023 (UTC)


 * I was often asked for an English introduction to this topic. At first I wanted to integrate it as a chapter to an already existing book, e.g. Statistics. But after I have checked how similar topics are written there, such as Statistics/Distributions/Normal_(Gaussian), I got the feeling that my text is much closer to what I expect from a textbook on a specific topic, with an introduction, detailed didactic explanations and application examples. It is actually much more than a Wikipedia article. After I could not find an suitable existing book, I finally decided to make it a stand-alone book here. But what makes a book a book? Should it become just bigger? Or part of a comprehensive superordinate set of topics? I don't see any comparable wikibook on statistics or machine learning that seems to develop towards the the same level of depth and didactics where I could integrate this as a chapter. Although this "small book" does not fulfill the goal of wikibooks for comprehensive big textbooks, it still fulfills the educational purpose of wikibooks and many people will appreciate the introduction to this complex advanced mathematical topic.--Physikinger (discuss • contribs) 08:32, 22 March 2023 (UTC)
 * @Physikinger Based on your description, it seems like it could be turned into a book. Based on policy, other books hosted here, and my own thoughts, I personally would recommend a few things to start making it less encyclopedic and more instructional:
 * Reformat it into multiple pages, with each one being a distinct chapter
 * Include a chapter with introduction, scope, and goals of the book
 * Include clear applications
 * Include questions and practice problems
 * Definitely reach out if you need help with any of this! —Kittycataclysm (discuss • contribs) 12:13, 22 March 2023 (UTC)
 * Some comments to your points:
 * "Reformat it into multiple pages, with each one being a distinct chapter"
 * This would not support the readability, since the size is actually not so huge. I see also the drawback in Wikibooks in general that there is not even a link at the end of each chapter to the beginning of the next chapter. This interrupts the flow of reading. I also don't think that the book will grow much further. It's a quite specific topic. I wrote 10 years on the article and got almost no help by other authors. I would more prefer to make it in future part of a book on machine learning, as a chapter.
 * "Include a chapter with introduction, scope, and goals of the book"
 * Ok, I can do that.
 * "Include clear applications"
 * There are already 3 applications at the end. One is a real application I used in industry.
 * "Include questions and practice problems"
 * There is already a free scientific textbook about Gaussian Processes which is "the bible" for this topic. There are more applications in this book. I don't want to replace or repeat this book. It should be just a more gentle introduction and practical guide.
 * Physikinger (discuss • contribs) 23:02, 22 March 2023 (UTC)
 * @Kittycataclysm I can't see a clear strategy behind the suggested split. What is the purpose? For a page of this size scolling with a scrollbar seems to me more convenient than an annoying click track. Physikinger (discuss • contribs) 21:06, 27 March 2024 (UTC)
 * Hi @Physikinger. I would like to split it for accessibility reasons. For some people, it is more accessible to navigate and edit by chapter, and for some it is easier to work with one big scroll. We have the printable template option that combines separate chapters into one long scroll, but the reverse is not possible. Cheers —Kittycataclysm (discuss • contribs) 22:26, 27 March 2024 (UTC)
 * There is an auto-generated table of content where you can click to jump to each chapter. You can also edit each section separately. The whole text is actually not so big that there would be an avantage of splitting. There are also several cross-links inside the document e.g. to examples or to the operations. I don't know if this would work with the printable template. There is also not even an auto-generated link at the end of each chapter to the next section. You would have to go back to the first page with the main table of contents in order to go to the next section. This would not feel like a book you can read continuosly from the beginning to the end. So I would rather keep it together for accessibility reasons. Physikinger (discuss • contribs) 00:23, 28 March 2024 (UTC)