Neural nets: How should one tweak the input slightly to increase the output?

A good example of how an idea expressed with too abstract a manner confuses me while giving me a visual generally makes it easy. Here is an early point in understanding what neural nets do, which I read elsewhere several times before seeing it here in a manner that makes it easy. Click on the image below to go to the original web page, Hacker's guide to Neural Networks, which is a lot easier for me to understand than the similar material in chapter 1 of the Neural Networks and Deep Learning book I mentioned earlier.

 

 

Posted in Neural Nets and AI Stuff on May 26, 2017