Today we recommend both an open source project and a book, written by Abhishek Thakur, a technology god who can help you solve (almost) all machine learning problems, and the open source project has a 6K star on GitHub, which is: approachingalmost.
What is approachingalmost?
approachingmost is the abbreviation of the author's project name, and the title of the book is: "Approaching (almost) any machine learning problem". Sounds like a nod to it, and this book explains how to deal with the challenges of ML and DL rather than simply explaining algorithms. The book contains a lot of code, which is suitable for readers who have a certain theoretical foundation in ML and DL and want to delve into the application of machine learning.
About the author
The author, Abhishek Thackur, is really not small, he has accumulated more than 1,000 medals on kaggle, and the highest ranking is third. He's the chief data scientist at boost.ai company, so he's really been working in the machine learning space for a long time.
In 2017, he published an article on Linkedin called Approaching (Almost) Any Machine Learning Problem, introducing an automated machine learning framework he built that can solve almost any machine learning problem, and this article was popular in Kaggle.
Later, the author enriched the content of his article as a whole, and finally published a book, which can be regarded as a high-quality introductory learning material for machine learning.
About books
"Approaching (almost) any machine learning problem" is a full 300 pages, which can be said to be very substantial. The author has won so many medals on Kaggle, and this book is arguably the best product of his theory + practice. The discussion in this book will focus more on the application of machine learning models, such as preprocessing steps. Below is the table of contents of the book
directory
- Setting up your working environment
- Supervised vs unsupervised learning
- Cross-validation
- Evaluation metrics
- Arranging machine learning projects
- Approaching categorical variables
- Feature engineering
- Feature selection
- Hyperparameter optimization
- Approaching image classification & segmentation
- Approaching text classification/regression
- Approaching ensembling and stacking
- Approaching reproducible code & model serving
In addition, most of the books have been translated by Internet enthusiasts, and some of the contents are relatively simple, but he has not translated them all, and the translated content is as follows:
After all, not every student's English is so good, so the translated version is really a good news for everyone, and I am very grateful to the students who translated it.
Project information
- 项目名称：Approaching (almost) any machine learning problem
- GitHub link: https://github.com/abhishekkrthakur/approachingalmost/tree/master
- Number of stars: 6K+
In order to facilitate students who are inconvenient to access GitHub, I have sorted out the original English version and the translated version, and if necessary, you can directly reply to AAAMLP in private message, and you can get the download.