This week I was going to upload a story about the importance of correctly coding categorical variables, but given that we are very close to Christmas, I thought it would be a good idea to drop some rules about books. discovered in the course of the year and that I myself would like to receive as a gift.
The recommendations will be divided into the following categories:
- Non-technical books
- Technical books
- Data visualization books
You will find 3 of each, and hopefully I'll surprise you with some.
Naked Statistics – Rating: 4 out of 5
Although some chapters may be too basic, Charles Wheelan guides us through many of the things that we usually use and mention in our daily lives, with classic mistakes and good practices around all those things concepts. Also interesting for getting some real-life examples of technical concepts if you are studying and not yet working in the field.
A bit more:
once considered annoying, the field of statistics is rapidly evolving into a discipline that Hal Varian, chief economist at Google, actually calls “sexy.” From batting averages and political polls to game shows and medical research, the real-world application of statistics continues to grow by leaps and bounds. How can we catch schools that cheat on standardized tests? How does Netflix know which films you like? What is causing the increasing incidence of autism? As bestselling author Charles Wheelan shows us in Naked Statistics, the right data and a few well-chosen statistical tools can help us answer these questions and more.
Algorithms to Live By – Rating: 4.5 out of 5
A bit slow in the beginning, and maybe even a bit repetitive sometimes, but with some really amazing chapters after that, this is a book you'll probably have to read a few times to fully process. In my case, it gave me tools for my daily life and work, as well as questions and intrigues about things that I continued reading after completing the book.
A bit more:
all our lives are limited by limited space and time, boundaries that give rise to a certain set of problems. What should we do, or undo, in a day or a life? How much junk do we have to accept? Which balance between new activities and known favorites is the most satisfying? These may seem unique human dilemmas, but they are not: computers also have the same limitations, so computer scientists have been struggling with their version of such problems for decades. And the solutions they have found have a lot to teach us.
Invisible Women: Data Bias in a World Designed for Men – Rating: 5 out of 5
As a friend of mine said to me: don't be put off because it's “about women” because that actually means it's men! I am not done with this book yet, but so far it has turned out great. Caroline Criado-Pérez's research is really impressive and reveals real facts about a problem that we should all be dealing with.
A bit more:
Imagine a world where your phone is too big for your hand, where your doctor prescribes a medication that is wrong for your body, where you are 47% more likely to be seriously injured in a car accident, countless hours of work you do not recognized or appreciated. If this sounds familiar, chances are that you are a woman.
Invisible Women shows us how we systematically ignore half the population in a world largely built for and by men. It exposes the data gap between men and women – a gap in our knowledge that underlies continuous, systemic discrimination against women, and which has created a profound but invisible preference with a profound effect on women's lives.
Bonus track: in case you don't want to read the whole book, or maybe you just want to take a look, another friend of mine has recommended a episode of the 99% invisible podcast where she interviewed the author.
Deep Learning with Python – Rating: 4 out of 5
Love Python? Do you want to know more about deep learning? Francois Chollet is the author of Keras, one of the most used libraries. This book ranges from the intuition behind it to the practical application of concepts, showing examples and making things enjoyable for the reader.
A bit more:
deep learning applies to an increasing number of artificial intelligence problems, such as image classification, speech recognition, text classification, answer to questions, text-to-speech and optical character recognition. It is the technology behind photo and tagging systems on Facebook and Google, self-driving cars, speech recognition systems on your smartphone and much more. Deep learning particularly excels in solving problems with the perception of machines: insight into the content of image data, video data or sound data. Here is a simple example: suppose you have a large collection of images and you want to link tags to each image, for example “dog “, “cat “, etc. Through in-depth learning you can create a system that understands how assign such tags to images and only learn from examples. This system can then be applied to new images, thereby automating photo tagging. A thorough learning model only needs to be fed with examples of a task to generate useful results on new data.
One hundred pages of machine learning book – score 5 out of 5
I started reading this book online, every time I needed to understand a concept better, but found it so good, clear and helpful that recently decided to buy the printed edition to have it at home. It goes through the most important concepts and algorithms in Machine Learning. Explain how they work in a technical and non-technical way.
A bit more:
supervised and uncontrolled learning, support of vector machines, neural networks, ensemble methods, gradient drop, cluster analysis and dimensionality reduction, autocoders and transfer learning, feature engineering and hyperparameter matching! Maths, intuition, illustrations, everything in just a hundred pages!
Bonus track: the book has its own web page , where you can find different layout options, as well as reviews and comments from different people.
Probability: for the enthusiastic beginner – Rating: 4 out of 5
A little rusty about your probability qualities and skills? Then this is the book for you. It goes through all important topics in a very friendly way, full of examples and elaborated problems. Review concepts such as combinatorics, the probability rules, Bayes' theorem, expectation value, variance, probability density, common distributions, the law of large numbers, the central limit theorem, correlation and regression. You don't need to know much more about this book. A handy book to have in your area at home.
Books about data visualization
I found data visualization such an important part of Data Science that I published a story some time ago with the title: '10 tips to improve your plot. Because plotting matters in practice “. In that publication I wrote a bit about why I think data visualization is so important, but if you are interested in reading more about it and improving your understanding and skills, the following books will certainly help you.
Good graphs – Rating: 4.5 out of 5
As the Hundred Page Book Machine book, I first found this book online, read a number of chapters, and decided to buy it to have at home. A great addition to any library, not only because it is a beautifully designed book, but also because it is a very pleasant book to read.
A bit more:
Good visualization can communicate the nature and potential impact of information and ideas more powerfully than any other form of communication. (…) Building good graphs is rapidly becoming a necessity for managers. If you don't do it, other managers will do it and be noticed and given the honor to contribute to the success of your business. In Good Charts, dataviz maven Scott Berinato provides an essential guide to how visualization works and how this new language can be used to impress and persuade. Dataviz today is where spreadsheets and word processors were in the early 1980s – to the point of changing the way we work. Berinato offers a system for visual thinking and better graphics through a process of talking, sketching and prototyping.
The visual representation of quantitative information – Rating: 4 out of 5
Not as attractive to the eye as good graphics, but extremely rich in content. This book was first published in 1983 and has been a sort of bible of data visualization for some time. While it is true that some things in the book are nowadays outdated or outdated, concepts and tools that you probably use today have been published for the first time.
A bit more:
the classic book about statistical images, graphs, tables. Data theory theory and practice, 250 images of the best (and some of the worst) statistical images, with detailed analysis of how data is displayed for accurate, effective, fast analysis. Design of the screens with high resolution, small multiples. Edit and improve images. The data-ink ratio. Time series, relational images, data cards, multivariate designs. Detection of graphic deception: design variation versus data variation. Sources of deception. Aesthetics and graphic data display.
Information is nice – Rating: 4.5 out of 5
This is a recommendation from a friend of mine, so although I can't give you my opinion about the book, I leave it behind full description below:
Every day, every hour, every minute we are bombarded with information, from television, from newspapers, from the internet, we are steeped in it. We need a way to relate to it. Enter David McCandless and his stunning infographics, simple, elegant ways to communicate with information that is too complex or too abstract to contain in a way other than visual. McCandless creates visually stunning displays that combine the facts with their connections, contexts and relationships, making information meaningful, entertaining and beautiful. And his genius is just as good at finding new ways to combine data sets provocatively as in finding new ways to show the results.
Knowledge is Beautiful is a fascinating twist through the world of visualized data, all of which bear the hallmark of the groundbreaking, distinctive style of David McCandless. The fascinating successor to the bestseller The Visual Diversum, Knowledge is Beautiful, offers a deeper, more comprehensive view of the world and its history, with more connectivity between the pages, a greater exploration of causes and consequences, and a more inclusive global outlook. Because some of the content comes from McCandless's international following, Knowledge is Beautiful achieves a revolutionary and democratic view of the most important issues of history and politics, the facts of science, literature flows and much more.
Well, I think that's enough, at least for Christmas. If you buy and read one, or if you have already done it in the past, leave me your own review. I would love to hear your thoughts!
And if you liked this story, check out a few of mine, such as 6 amateur mistakes I made working with train test splits or scraping in 5 minutes . All available in my profile .AI, algoritme, boeken, kerst, Kunstmatige intelligentie, lezen, naked statistics, studeren