I'm looking books on Introductory Statistics and Basic Experimental Statistics (Designs) for Computer Scientists. I'd highly appreciate if you point out few. Thanks


Actually I've been given the responsibility to teach:

  • Stat-700 Elements of Statistics & Biometry (Basic Descriptive and Inferential Statistics)
  • Stat-701 Experimental Designs (CRD, RCBD, Latin Square, Factorial Experiments, Split Plot, Regression)

courses to Computer Science Graduate Students. These two courses have their origins from Agriculture but now these courses are also mandatory for Computer Science Graduate Students. Rather than using examples from Agriculture, I would prefer to discuss examples from Computer Science.


2 Answers 2


There are two ways I can see making these things interesting to graduate students in computer science. You can share statistical problems with them and ask them to think critically about the computational problems. Or you can share computational problems with them, and ask them to think critically about how statistics can help them reason about those problems.

By this time they're very familiar with algorithms, and comfortable receiving new computational problems. You could take a look at these course notes for inspiration on giving them statistical problems to reason about with their computational skills. By implementing their own algorithms to compute the statistics, they'll get some intuition about what the results mean.

On the other side of the same coin, they can get an appreciation for why these statistics matter to them by learning about applications of statistics in computer science, which I think is the main thrust of your question. To that end, there's plenty to draw from. Things in this category include voice and image recognition, recommendation systems, cryptography, optimization and parallelization.


There are already quite some questions like this on this site, with good answers. Some are What should a graduate course in experimental design cover?
Recommended Text for Essential Experimental Statistics or search this site!

But you say "for computer scientists". What will you do use experimental design for? One specific thing of interest for computer science is design for computer experiments. They are different from the usual kind in that the outcomes are not really random, but deterministic (if we talk about "simulations" with deterministic models like complicated computer codes for differential equations). Then, the usual ideas behind factorial experiments do not really apply, and you need some kind of "space-filling designs".

A book going into such ideas (like latin hypercubes) is "Design and Modeling for Computer Experiments" by Kai-Tai Fang, Runze Li and Agus Sudjianto: http://www.amazon.com/Modeling-Computer-Experiments-Chapman-Analysis/dp/1584885467/ref=sr_1_2?s=books&ie=UTF8&qid=1455574142&sr=1-2&keywords=experimental+design+for+computer+experiments


You can find a lot of information by googling, for instance the search term "use of experimental design in computer science". Here http://sing.stanford.edu/cs303-sp11/ is syllabus for a course similar to the one you have to teach (701). Start with possible uses for experimental design within computer science:

   Comparison of algorithms
   Comparison of user interface designs
   planning of marketing campains
  • $\begingroup$ Thanks for your answer and please have a look on my edits. $\endgroup$
    – MYaseen208
    Commented Feb 16, 2016 at 2:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.