Source: Deep Learning on Medium
In what scenario perceptron model is ahead from mp neuron model?
Data: mp model receives inputs in boolean forms only, but the perceptron model accepts data in terms of real inputs. But outputs are boolean in both models
Task: mp model will give boolean outputs, but the perceptron model will give as classification form.
model: mp neuron model is linear(for every input same slope), but the perceptron model has weights for every input(different slopes for different inputs)
loss function: not much difference loss.
learning algorithm: mp model use brute force search(manual search), but in case of the perceptron model algorithm do the job
data preparation in the perceptron model:
perceptron model will take real inputs, but it follows the standardization.
What, weights allow us to do?
based on is weight(w) is positive or negative and on value, we can control output(is the phone is liked?) for those specifications (x).
perceptron math fundae