It is very interesting to read a hard core computing book from authors those are not in the computing industry. Their area of studies really make them as right choice for writing this book. Drew Conway studies international relations, conflicts and terrorism using tools of mathematics, statistics and computer science. John Myles White in “how humans make decisions”. Obviously, both are Ph.D students.
Machine learning is one of the grooming area in the computing world which is actually a branch of artificial intelligence. Based on the given data, we can capture characteristics of interest of our unknown underlying probability distribution. This book thoroughly covers various disciplines such as classification, ranking, regression, regularization, optimization, etc with practical examples using “R” language.
If you are a programmer, you may little bit astonished first time to use R. Later, you will understand that “R” is not the language of programming, but excellent companion for people in statistical field. So, you may not fully comfortable with “R” after reading the first chapter “Using R”. Authors also mentioned that “R remains a relatively niche language, even among experienced programmers”. However, It would be good if they explain R syntax and its usage crisp and short in the Appendix section.
The chapter 1 (Using R) and 2 (Data Exploration) make you start the journey slowly with many good theories on data analysis. The visual explanation of “data as rectangle”, MxN matrix of data into single row or single column view are good learning. Inferring data is another good point. Wherever required,visual representations come for you to understand. Otherwise, they comfortably explains the concepts textually. Agile people should be patience on this.
“Chapter 3Classification : Spam filtering” make your journey at the highest speed, with lot of interesting turns. After two chapters with academical effect, this chapter make you feel like reading fictions. They starts with an example of how can you predict a person is man or woman based on the weight and height. They called this mechanism as “separating hyperplane” and also explained a way of taking decision called “kernel trick”.