Data Mining Book Review: Data Science for Business
Foster Provost and Tom Fawcett are known for their work on fraud detection, among others. I have recently read their last book, Data Science for Business – What you need to know about data mining and data-analytic thinking. No suspense: it’s one of the best data mining book I have ever read. Its style allow it to be read by beginners, but its wide coverage and detailed case studies makes it a reference for experts as well.
As the title suggest, the book has a real focus on business with plenty of industry examples and challenges. The style is very pleasant since authors have made efforts to put the reader in specific situations to better understand a problem. To be noted the very interesting discussion of data mining leaks as well as data mining automation. The book is divided by concepts and provides a focus on them (instead of techniques). Although no exercice is present, the book could easily be used as a resource for a course.
Each chapter is clearly divided into basic and advanced topics. The evaluation phase of the data mining standard process is deeply discussed. The section about Bayes rule is very well written. Data Science for Business is also an excellent resource to avoid data mining pitfalls. Chapter 13 is a must-read in order to understand success factor for implementing data mining in a company. To conclude, targeted at both beginners and experts, Data Science for Business is the new reference for data mining professionals working in industry.