Re: [Gneuralnetwork] mathematical background

[Jean Michel Sellier]

Well, to make a very long story short: the basic theory on which neural networks are based is exactly the same since a long time so, if you are looking for a good and understandable introduction to the subject, this is definitely a good book. For more recent stuff, you can always surf the internet and/or use some other book once you are familiar with the basic knowledge ;)

I hope this helps!

JM

2016-03-23 19:58 GMT+01:00 Ivan F. V. B. <address@hidden>:

* Jean Michel Sellier <address@hidden> [2016-03-23 19:05]:
> I agree with the list of mathematical knowledge required that you
> reported
>
> Personally, I would recommend the book of Bishop, "Neural Networks for
> Pattern Recognition".

Thanks Jean for your reply.

I remember that title as the bible in some circles, but that was like 20
years ago. Haven't a lot happend in that time, like deep learning
accomplishments, self driving cars and robotics?

Ivan F. V. B.

> 2016-03-23 18:41 GMT+01:00 Ivan F. V. B. <address@hidden>:
>
> > Dear GNeuralNetworkers,
> >
> > it is very exiting to have heard of this community and start to be part
> > of it.
> >
> > Unfortunately, I must admit that my mathematical background is limited and
> > rusted. Thus my question:
> >
> > Which mathematical fields would you recommend to revise/learn and to
> > which level of deepness?
> >
> > Would you agree or extend the syllabus of the Machine Learning Nanodegree
> > of
> > udacity.com, which I copied here for convenience from
> > https://www.udacity.com/course/machine-learning-engineer-nanodegree--nd009
> > ?
> >
> > - Intermediate statistical knowledge
> >
> >   -  Populations, samples
> >   -  Mean, median, mode
> >   -  Standard error
> >   -  Variation, standard deviations
> >   -  Normal distribution
> >   -  Precision and accuracy
> >
> > - Intermediate calculus and linear algebra
> >
> >   - Derivatives
> >   - Integrals
> >   - Series expansions
> >   - Matrix operations through eigenvectors and eigenvalues
> >
> > Would anyone by interested in co-writing free (libre) accompanying
> > materials
> > for understanding and using machine learning algorithms with
> > gneuralnetwork,
> > including cute examples?
> > Little cute projects like this have some traction
> > https://github.com/yenchenlin1994/DeepLearningFlappyBird
> >
> > A more extensive syllabus on math background can also be found in the
> > Introduction to Machine Learning - Cambridge University Press 2008
> > available online at http://alex.smola.org/drafts/thebook.pdf
> > but not under a free (libre) license.
> >
> > Do you know any other good resource on math background?

--

Jean Michel Sellier, Ph.D., M.Sc.

www.nano-archimedes.com

www.gnu.org/software/archimedes

www.gnu.org/software/nano-archimedes

www.gnu.org/software/gneuralnetwork

http://en.wikipedia.org/wiki/GNU_Archimedes