Consider the following neural network. Single-circled nodes denote variables (e.g. x is an input variable, h...

Transcribed Image Text:

Consider the following neural network. Single-circled nodes denote variables (e.g. x is an input variable, h is an intermediate variable, is the output variable), and double-circled nodes denote functions (e.g. takes the sum of its inputs, and denotes the logistic function (x)=1+- Suppose we have a loss L(y,y) = ||y - ||. We are given a data point (x1, x2, X3, X4) = (-0.3, 4.9, 1.1, -2.7) with true label 0.7. Use the backpropagation algorithm to compute the partial derivative 44 for all wi. -1.7 goood W = 0.1 0.6 good == -1.8 h2 = -0.2 83 W6 = 0.5 Consider the following neural network. Single-circled nodes denote variables (e.g. x is an input variable, h is an intermediate variable, is the output variable), and double-circled nodes denote functions (e.g. takes the sum of its inputs, and denotes the logistic function (x)=1+- Suppose we have a loss L(y,y) = ||y - ||. We are given a data point (x1, x2, X3, X4) = (-0.3, 4.9, 1.1, -2.7) with true label 0.7. Use the backpropagation algorithm to compute the partial derivative 44 for all wi. -1.7 goood W = 0.1 0.6 good == -1.8 h2 = -0.2 83 W6 = 0.5