Fog Creek Software
Discussion Board




engineering statistics book recommendation?

I have a math degree, but somehow, never took any real statistics courses. Can anyone recommend a good statistics book for self-study?

Preferably something that will help me understand the statistics used in engineering, rather than something oriented to say, market analysis or public health.

My goal is to be able to understand the papers on this page:

http://ic.arc.nasa.gov/projects/bayes-group/super-res/2d/

Thanks in advance!


Monday, January 19, 2004

"Introduction to Probability and Statistics Principles and Applications for Engineering and the Computing Sciences"
by Jesse Arnold et. al., ISBN 007246836X

an updated version of what I used in college.  Not the most readable statistics book I've ever used, but definately sound, and aimed at the types of topics that you're interested in.

Elephant
Monday, January 19, 2004

this is the text MIT uses:

Probability and Statistics (3rd Edition) by DeGroot and Schervish

http://www.amazon.com/exec/obidos/tg/detail/-/0201524880/qid=1074537626/sr=1-10/ref=sr_1_10/002-7118909-9286469?v=glance&s=books


would this be a good book for self study?


Monday, January 19, 2004

elephant, check out the first 3 amazon reviews for that text...hehe.

"This is the worst text I have ever used" (x2)
"This book made me hate statistics"


Monday, January 19, 2004

My fiance just finished teaching herself calculus based statistics for the Actuarial exams.  Although, not directly aimed at engineering principles, I will say that their recommended selection of text books, along with the Course 1 DVD review from ACTEX did the trick for her course in self study, as she passed the exam.  I've provided a list of the SOA's recommended texts below, and you can find more information at www.soa.org

A First Course in Probability (Sixth Edition), 2001, by Ross, S.M., Chapters 1-8.

Fundamentals of Probability (Second Edition), 1999, by Ghahramani, S., Chapters 1-10.

Probability for Risk Management, 1999, by Hassett, M. and Stewart, D., Chapters 1-11.

Probability and Statistical Inference (Sixth Edition), 2001, by Hogg, R.V. and Tanis, E.A.,
Chapters 1-6.

Probability: The Science of Uncertainty with Applications to Investments, Insurance and Engineering 2001, by Bean, M.A., Chapters 1-9.

Elephant
Monday, January 19, 2004

And in response to people hating the book, a lot may be students that used it in the Virginia Tech course.  The professor was an asshole and a lousy teacher.  It made me hate the book the first time around, however the second time around, it was quite good.  But if you're interested in self study, I'd probably shy away from it.  Maybe see if you can get it from the library or something to give you a good selection of CS and EE oriented problems.

Elephant
Monday, January 19, 2004

I've used several books on applied statictics and DOE written by Douglas C. Montgomery which were pretty good.

An Amazon listing of his books is here:
http://www.amazon.com/exec/obidos/search-handle-url/index%3Dstripbooks%26field-keywords%3Ddouglas%2520c.%2520montgomery/104-5222291-1771927

His books generally get good reviews, and I found that for the most part he was readable and the material was applicable to real-world scenarios.

If I had one complaint it is that sometimes there would be passages with the logic "If we know A and B, it can obviously be shown that D," leaving me scratching my head thinking there was a missing C and that nothing was "obvious." My point is that what is obvious to the author may noty be obvious to you.

Nick
Monday, January 19, 2004

Favourite statistical algorithms I have used #15

Double exponential smoothing


Tastes so good.

Simon Lucy
Monday, January 19, 2004

Blank's article gives this example:

http://ic.arc.nasa.gov/projects/bayes-group/super-res/2d/VO217R1.html

Check out the original Viking lander image and the one with improved resolution after statistical processing.

HOLY COW!

I want to learn statistics like that too.

ANY SUFFICIENTLY ADVANCED TECHNOLOGY IS INDISTINGUISHABLE FROM MAGIC.

Dennis Atkins
Monday, January 19, 2004

He combined 24 images to do that, it's not *that* magic.

www.MarkTAW.com
Monday, January 19, 2004

It would be easy if it was making a landscape by stitching together multiple x-res images to create a n-times bigger x-res image, but this is multiple x-res images to create a kx-res image where k is a surprisingly large number.

i'd like to see you try it if it's so easy! Read NASA's paper on the thing and then get back to me that it's not magic.

Dennis Atkins
Monday, January 19, 2004

Oh, I'm not saying it's not difficult, I just wanted to let anyone who didn't follow the link to know that it's not one image extrapolated into an amazingly high res image, but one put together from 24 seperate images.

www.MarkTAW.com
Monday, January 19, 2004

*  Recent Topics

*  Fog Creek Home