----of machine learning--the theoretical basis and proof of VC dimension

The VC theory proves that through a series of upper bounds, an upper bound formula for all objective functions and all training data sets is obtained, which is of great significance for machine learning. But it is also because as follows more upper limit, so this value to guide practice is only a kind of worst reference, there are too many hypothesis set cannot find VC Bellavita. It can be seen that the process of proving is very skillful, subtly transforming infinity into finite, and then finding bounds.

In the VC-dimensional theory to prove that involves the growth function, break-up and other important concepts, many friends in the understanding may be unintelligible, perhaps there is an error (including the author himself is also so). Through this study, we can firmly study the idea, he told us to learn through the sample data, and then applied to the data not seen is theoretical basis. This article is a lot of formulas, the meaning of understanding each step, the text of the inappropriate, please feedback with me.

Link: http://pan.baidu.com/s/1pLEI3On Password: gsn4

----of machine learning--the theoretical basis and proof of VC dimension