what is a KIV neural network?
Question by nani d: what is a KIV neural network?
IT belongs to neural network.
Answer by Don M
The so-called K model neural networks are biologically inspired models that conceive the network as consisting of layers corresponding to data gathering, perception, an intermediate level and on up. Neurons from each level may excite or inhibit other neurons, usually in neighboring levels. At the lowest (K0) level, acquisition of a signal is modeled as an asymmetric multidimensonal sigmoidal function (S-curve with multiple inputs that may arrive at any time) whose X axis represents some quality/ies of the signal and whose Y axis represents the likelihood that the overall signal will be treated as excitatory. If the signal passes that threshold, the K0 level neuron that received it may excite or inhibit another neuron one layer back. This layer (k1) is often called the perceptual layer, as it deals with the interpretation of the signal and with multiple-neuron behaviors such as “feature detection”. In biological systems, for instance, the K0 level neuron in the visual system may detect a black pixel, but the K1 level will interpret several black pixels in a row as a line (loosely speaking). Subsequent levels may deal with more complex interpretations such as movement, judging whether the signal should be attended to by higher layers, etc. At each layer, inhibition or excitation of neighboring neurons will affect the “weights” associated with the inputs on the neurons touched. Over time, the system learns by developing a set of weights (based on excitation / inhibition of the sigmoidal function in each neuron) that match the desired behavior.
k4 refers to the fact that it is a neural network with 4 years, usually composed of multiple lower-level networks.
At some level, most such neural networks also have an “instinct level” reward apparatus that allows the network to judge whether it has produced the desired behavior, and this is fed back through the network in a more-or-less reverse order as a means of reinforcement of the behaviors.
I know this is a fast-and-loose definition, but hopefully that helps.
Give your answer to this question below!
No related posts.