+ 2

What about Neural Network ??

5th Nov 2017, 5:57 PM
Raj kumar Singh
Raj kumar Singh - avatar
3 ответов
+ 10
Neural network is a setup of interconnected perceptrons ("neurons") that imitates the behaviour of the protein-based original -- the brain. Each of perceptrons contains an activation function which reacts on the input, transforms it by multiplying by a hidden weight and produces the output. The overall output of all neurons is consolidated at the end to produce the final output to make a decision (classification, regression, etc.). Of course, as with the real brain -- what it "thinks", might not be true. It is always based on probability and allows a degree of error. During the process of neural net learning the weights are initially randomized and then slowly fine-tuned until the aggregate error (the difference between the value predicted by the network and the true value) is the lowest possible (at least in theory).
5th Nov 2017, 6:13 PM
Kuba Siekierzyński
Kuba Siekierzyński - avatar
+ 7
What about it? :)
5th Nov 2017, 5:59 PM
Kuba Siekierzyński
Kuba Siekierzyński - avatar
+ 3
have any information about it please describe..
5th Nov 2017, 6:00 PM
Raj kumar Singh
Raj kumar Singh - avatar