Mailing List Archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [tlug] Open Access Journals

On Apr 4, 2014 11:25 AM, "Darren Cook" <> wrote:

> > Now, it's almost reached a point in bioinformatics where we don't know
> > what we're looking let's throw it into a piece of software
> > (i.e., a machine learning algorithm for example) -- it tells us which
> > genes out of the thousands that were examined are interesting in a
> > statistically significant way, and then we write a paper about it.
> There does seem to be a paradigm shift in the way science is done.

There have always been many models of  "the way science is done".
Consider a botanist working in a previously unexplored area, and a
medical researcher evaluating a newly identified blood agent as an
indicator for a type of cancer, and cosmologist busily working on the
ripples in space-time in the nanoseconds after the big bang. All
scientists, all working in totally different paradigms.

I relate to the quote above because my own research involves using
AI techniques to detect useful patterns in large amounts of text. It's
a fairly foreign concept to people brought up on traditional views of
the "scientific method", and I confess to a little irritation with some
practitioners in NLP who chew over a lot of very noisy data and then
come to conclusions based on results that aren't within a bull's roar
of being statistically significant (I majored in stats way back in the
60s.) Still "social scientists" have been getting away with that for 


Home | Main Index | Thread Index

Home Page Mailing List Linux and Japan TLUG Members Links