David MacKay, the late Cambridge physicist, wrote a free textbook on information theory, inference and learning algorithms, and made the Dasher text-entry tool.
Commenters call his book excellent and the best resource for the math behind modern machine learning, and also recommend his YouTube lectures.
> New solar cells can sometimes take greater than the life of the product to recoup the energy spent on creating them. If you live in, say, Massachusetts [...] you'll probably never net an energy savings.
"False. The energy yield ratio (the ratio of energy delivered by a system
over its lifetime, to the energy required to make it) of a roof-mounted,
grid-connected solar system in Central Northern Europe is 4, for a system
with a lifetime of 20 years (Richards and Watt, 2007); and more than 7 in a sunnier spot such as Australia. (An energy yield ratio bigger than one
means that a system is A Good Thing, energy-wise.) Wind turbines with a
lifetime of 20 years have an energy yield ratio of 80."
source:
http://www.inference.phy.cam.ac.uk/withouthotair/c6/page_41.shtml
(I assumed that the sun in Central Northern Europe shines with an intensity similar to the intensity in Massachusetts.)
Do windmills kill “huge numbers” of birds? Wind farms recently got adverse publicity from Norway, where the wind turbines on Smola, a set of islands off the north-west coast, killed 9 white-tailed eagles in 10 months.
I share the concern of BirdLife International for the welfare of rare birds. But I think, as always, it’s important to do the numbers. It’s been estimated that 30 000 birds per year are killed by wind turbines in Denmark,
where windmills generate 9% of the electricity. Horror! Ban windmills! We also learn, moreover, that traffic kills one million birds per year in Denmark. Thirty-times-greater horror! Thirty-times-greater incentive to ban cars! And in Britain, 55 million birds per year are killed by cats.
--David MacKay, "Sustainable Energy -- without the hot air".
http://www.inference.phy.cam.ac.uk/withouthotair/c10/page_63.shtml
> Thus solar and wind fit in with our natural rhythm quite nicely.
It's not that nice, because power output from wind/solar plants is random (depends on things like cloud cover or wind strength) and it causes problems in power grid, where power demand must meet the supply exactly, or bad things will happen.
"Without Hot Air"[1] covers this, and other renewable-related issues quite nicely and with real data. I recommend it, it's a good read. It has some good ideas on how to solve power supply/demand problems.
On TED2012 there was a talk about a new kind of batteries designed to solve those kind of problems in the power grid; the video from the talk is not yet up, though (but I think it should be soon, at [2]).
[1] - http://www.inference.phy.cam.ac.uk/withouthotair/
[2] - http://www.ted.com/talks?lang=en&event=2012&duration=&sort=newest&tag=
some useful books available as a PDF:
David Barber "Bayesian Reasoning and Machine Learning"
http://web4.cs.ucl.ac.uk/staff/D.Barber/pmwiki/pmwiki.php?n=Brml.HomePage
-- Thorough and systematic, accompanied by a Matlab library
Marc Steyvers "Computational Statistics with Matlab"
http://psiexp.ss.uci.edu/research/teachingP205C/205C.pdf
-- Somewhat unfinished but very good hands-on guide for a total novice
David Draper "Bayesian Modeling, Inference and Prediction"
https://users.soe.ucsc.edu/~draper/draper-BMIP-dec2005.pdf
-- Quite good if you're already familiar with a field, might be slightly daunting for a beginner
and the classic, David MacKay's
"Information Theory, Inference, and Learning Algorithms"
http://www.inference.phy.cam.ac.uk/itprnn/book.html
i'm not going to argue here because http://www.withouthotair.com/ makes the point better than i ever could.
see, for example, http://www.inference.phy.cam.ac.uk/withouthotair/c19/page_114.shtml - Don’t be distracted
by the myth that “every little helps.” If everyone does a little, we’ll
achieve only a little. We must do a lot. What’s required are big changes in
demand and in supply.
and, of course, we may be talking at cross purposes - there is some middle ground where a large-ish number of small-ish things can help. i am not arguing against maths, but i am saying that our natural assessment of what is significant without actually doing the maths is often misled by absolute values, as here.
anyway, if you haven't read that book, i recommend it.
As a rather short (9 pages) introduction to Bayesian inference, I remember greatly appreciating an extract from the book "Information Theory, Inference, and Learning Algorithms" by David MacKay.
The relevant part is on PDF pages 469 to 477 here: <http://www.inference.phy.cam.ac.uk/itprnn/book.pdf>
It explains how Bayesian inference works (which is intuitive and not strange at all) and gives examples of absurdities that comes upon you when you don't do it this way.
Highlights:
• "Let me through, I’m a Bayesian" (When analysing the effectivness of a vaccine)
• "I have no problem with the idea that there is only one answer to a well-posed problem" (In response to sampling theorists wide selection of ad hoc procedures)