I would like to interpret problem on picture below, which is about perceptron learning. It is about supervised learning wiht a training set, so correctness of values should be checked against a predefined set of values. I don't exactly know, how A, B
and bias(b)
values come. Could you please explain meaning of these and how these computed and changed during the learning process?
Asked
Active
Viewed 220 times
-2
![](../../users/profiles/7003167.webp)
appkovacs
- 97
- 1
- 9
1 Answers
0
Here you have an intuitive, visual, interactive and beautiful guide to the basics concepts of neural networks: https://jalammar.github.io/visual-interactive-guide-basics-neural-networks/
It will problably solve all your doubts. However, if you still have more questions after the reading, you will be able to ask something more specific. Enjoy!
![](../../users/profiles/3782161.webp)
Pablo EM
- 4,960
- 3
- 21
- 35
-
That is great, thanks for the link, but I would like to do these on a paper, manually. – appkovacs Dec 16 '16 at 11:44
-
Welcome! Yes, the examples are quite simple, you should be able to reproduce it manually, on a paper. After that, for sure you can face with your example. Or what do you mean exactly? – Pablo EM Dec 16 '16 at 11:49
-
Yes, I see, that there are 2 weights in my example, but don't know how to modify bias after calculating Y value. – appkovacs Dec 16 '16 at 12:10
-
Yeah, first you have to read carefully (and understand) the information in the link, then you will know how to modify the bias. – Pablo EM Dec 16 '16 at 14:04