Why is there nn at the beginning

And in the beginning there was the algorithm

The human brain as a model

(Artificial) neural networks (NN) represent a further branch of AI in addition to the ML. NN recognize patterns in data sets and assign the appropriate data. This allows models to be created and logical summaries to be formed. A network of interconnected neurons simulates the structure and processes in the human brain.

The neurons work in small data processing units. They are arranged in layers, connecting channels forward the information data. So that the individual neuron can react to this, a weighting and a threshold value must be set. The weighting is specified in the connections and determines how great the influence of the data is on the neuron: a positive weight has a reinforcing influence on another neuron, a negative weight has an inhibitory effect. The threshold determines the weighting from which the neuron is activated. It all happens in the first layer: the input layer.

The information data that activate the neurons through the connection channels are passed on to the next layer: the activation layer. The more activation layers there are, the more precisely the NN can filter and organize information. Once everything that is unimportant has been filtered out, the important information is output in the last shift.

The industry uses NN, for example, in quality control, in robot control or in capacity planning. In marketing, neural networks help to determine target groups and to carry out consumption analyzes. NN support AI applications in file management and speech recognition. In public life, for example, they can optimize timetables and traffic light switching. And last but not least, NN have already met everyone as the "brain" behind Alexa, Siri and Co.