Highly Recommended Books for Machine Learning Researchers

Frankly speaking, there are many recommendations out there (e.g. see this thread and related threads on Quora). But, hoping to blend together academic and industry concerns, I would like to provide here recommendations for machine learning researchers made by only two people: Michael Jordan, a distinguished professor from University of Berkeley, and Bradford Cross, a founder of Prismatic.

M. Jordan's list

Recommended books as a background for machine learning

Michael Jordan suggests for reading books on probability theory and statistics, optimization theory, information theory, and analysis.

The sources are as follows:

• a thread on HackerNews
• a thread on Reddit.

The full list of references is available here.

B. Cross' list

A couple of years ago, based on his experience, Bradford Cross gave a comprehensive list of the best resources on machine learning and the prerequisites in his blog ("Measuring measures"). According to him, these basics are essential:

• Analysis
• Linear Algebra
• Probability
• Statistics and nonparametric statistics
• Optimization
• Machine learning
• Feature Selection.