Output/visible layer

The two main categories of NN architectures are feedforward and feedback networks. Feedforward (or multilayer) networks consist of several interconnected layers I of neurons (Figure 9.1). The last layer L is called the output or visible layer, and the others are called the hidden layers. The number of neurons A, per hidden layer depends on the problem considered and is usually specified by trial and error that is, the more difficult the problem, the larger the required sizes of the hidden layers. Feedback networks consist of N neurons, with the output of each individual neuron being fed back to all others via weights w j. The operation of this network is given in terms of defined dynamics, which describes the time evolution of the neuron outputs. 

Figure6.25 Schematicdrawingofan artificial neural network with a multilayer perceptron topology, showing the pathways from the input Xj to the output y , and the visible and hidden node layers.

Type II supernovae are massive stars, ones that progress in their nuclear fuels well past the fusion of carbon and the fusion of oxygen at their centers. When their cores run out of nuclear fuel, those central regions collapse to form a neutron star, or in some cases a black hole. The incredibly intense emission of neutrinos from the newly born neutron star so heats the overlying layers, aided by an outward moving shock wave of pressure, that those layers pardy explode and are ejected. The last of these thatwas visible to the naked eye occurred in 1987, and demonstrated for the first time the correctness of the intense neutrino burst that is their main energy output. 

---