Recommended books by the UC Berkeley SGSA.
Well-written, go-to reference for all things involving categorical data.
Theoretical take on GLMs. Does not have a lot of concrete data examples.
Undergraduate-level textbook, has been used previously as a textbook for Stat 151A. Appropriate for beginners to R who would like to learn how to use linear models in practice. Does not cover GLMs.
Classic, approachable text, [Available online.]
Comprehensive but superficial coverage of all modern machine learning techniques for handling data. Introduces PCA, EM algorithm, k-means/hierarchical clustering, boosting, classification and regression trees, random forest, neural networks, etc. ...the list goes on. [Available online.]
The primary text for Stat 210A. [Download from SpringerLink.]
A good reference for Stat 210A.
Some students find this helpful to supplement the material in 210B.
What the majority of Berkeley undergraduates use to learn probability.
This text is more mathematically inclined than Pitman's, and more concise, but not as good at teaching probabilistic thinking.
What students in EECS use to learn about randomized algorithms and applied probability.
This is the standard text for learning measure theoretic probability. Its style of presentation can be confusing at times, but the aim is to present the material in a manner that emphasizes understanding rather than mathematical clarity. It has become the standard text in Stat 205A and Stat 205B for good reason. [Available online.]
This epic tome is the ultimate research level reference for fundamental probability. It starts from scratch, building up the appropriate measure theory and then going through all the material found in 205A and 205B before powering on through to stochastic calculus and a variety of other specialized topics. The author put much effort into making every proof as concise as possible, and thus the reader must put in a similar amount of effort to understand the proofs. This might sound daunting, but the rewards are great. This book has sometimes been used as the text for 205A.
This text is often a useful supplement for students taking 205 who have not previously done measure theory.
This delightful and entertaining book is the fastest way to learn measure theoretic probability, but far from the most thorough. A great way to learn the essentials.
Stochastic Calculus is an advanced topic that interested students can learn by themselves or in a reading group. There are three classic texts:
These are indispensable tools of probability. Some nice references are
Starting with elementary examples, this book gives very good hints on how to think about Markov Chains.
A theoretical perspective on this important topic in stochastic processes. The text uses Brownian motion as the motivating example.
Second book is more advanced than the first. Everything you need to know about matrix analysis.
A great book for self study and reference. It starts with the basis of convex analysis, then moves on to duality, Krein-Millman theorem, duality, concentration of measure, ellipsoid method and ends with Minkowski bodies, lattices and integer programming. Fairly theoretical and has many fun exercises.
Nice and easy to digest. Good as companion for 205A.
There's also a course on combinatorics this semester in the math department called Math 249: Algebraic Combinatorics. Despite the scary "algebraic" prefix it's really fun. [Available online.]
Great overview of sequencing technology for the unacquainted.
Great R code examples from computational biology. Discusses the basics, such as the greedy algorithm, etc.
This book is a good overview of numerical computation methods for everything you'd need to know about implementing most computational methods you'll run into in statistics. It is filled with pseudo-code but does use Maple as it's exemplary language sometimes. It has been a great resource for the Computational Statistics courses (243/244). Depending on what happens with this course, this may be a good place to look when you're lost in computation.
MIT OpenCourseWare 6.046J / 18.410J ''Introduction to Algorithms'' (SMA 5503) was taught by one of the authors, Prof. Charles Leiserson, in 2005. This is an undergraduate course and this book was used as the textbook.