Mastering Pattern Recognition: A Guide to Intuitive Learning

[Avatrin]

I am currently trying to learn pattern recognition from Pattern Classification by Duda et al. However, this books is a bit too dense for me. I keep hitting walls while trying to read the books (I understand most of the reasoning, but it just becomes too much and too abstract).

Can anybody recommend a book that eases you in more? Or, gives more intuitive explanations of the results? I am trying to get through the first five chapters of the second edition of the book.

 

[jedishrfu]

How about this book by Bishop?

https://www.amazon.com/dp/0387310738/?tag=pfamazon01-20

where some of the book's code written in MATLAB has been archived on github:

https://github.com/PRML/PRMLT

 

[StatGuy2000]

I would stand by jedishrfu's recommendation of Bishop's book.

Here are some additional suggestions:

Pattern Recognition and Neural Networks, by Ripley

https://www.amazon.com/dp/0521460867/?tag=pfamazon01-20

Elements of Statistical Learning, by Hastie et al.

https://www.amazon.com/dp/0387848576/?tag=pfamazon01-20

(the R functions and the additional resources for the above link can be found in the link below)

http://statweb.stanford.edu/~tibs/ElemStatLearn/index.html

 

[Chronos]

So, I gather you have mastered 'wall' pattern recognition ; Seriously, though, pattern recognition is a complex skill that humans master rather easily. Even birds do not seem to 'recognize' images in the clouds much less those of: dragons, angels,ships, or faces of their fledgeling friends. It took millenia of adaptations for us to acquire the algorithms and memory cross wiring to facilitate this ability, endowing each individual with unique pattern recognition skills. Programmers faces an enormous challenge in writing the alogithms and reference libraries necessary to replicate this ability.