now I am studying about an Artificial Neuron Network (ANN)
so an ann is a layered network of artificial neurons the common properties are:
- directed graph (with or without cycles)
- nodes are(primitive) computing devices(call neurons)
- edges carry (weighted) information - called weights
Inside of a neural network
- the interconnection pattern of the neurons or the ANN architecture
- The learning process for updating weights of the interconnection
- The activation function of each neuron that converts a neuron's inputs into its outputs
Supervise learning
- Supply the network with inputs and the desired output the weight are modified
Unsupervised
- Only supply inputs
- The neural network adjusts its own weights so that similar inputs cause similar outputs
- The system is supposed to discovery statistically salient features of the input population
ไม่มีความคิดเห็น:
แสดงความคิดเห็น