Homework Help is Here – Start Your Trial Now!
Engineering
Computer Engineering
Backward Step (i.e. back propagation) and Gradient Descent When training a neural network model we use back propagation to update the weight/bias parameters using gradient descent (or some variation of it). Gradient descent utilizes the chain rule, since each layer of a neural network can be described by a function, and the activation function (and additional multiple layers) can be described by a composition of functions. Suppose that we have the following functions: fi (x, w, b) = w·x + f2 (:) = ReLU (·) f (:) = f2 (fi (•)) note: the dot is used a placeholder here, so you could assume it to be u, or x, or any variable) Assume that we are going to carry out back propagation and a gradient descent step. Suppose that Ne = 0.5, and that x = 0.9, and that the learning rate is equal to 0.01. You can assume that b > 0. What will be the value of wt+1? note: be sure to carry out your answer to at least two (2) decimal places, i.e. 1e-2 precision)

Backward Step (i.e. back propagation) and Gradient Descent When training a neural network model we use back propagation to update the weight/bias parameters using gradient descent (or some variation of it). Gradient descent utilizes the chain rule, since each layer of a neural network can be described by a function, and the activation function (and additional multiple layers) can be described by a composition of functions. Suppose that we have the following functions: fi (x, w, b) = w·x + f2 (:) = ReLU (·) f (:) = f2 (fi (•)) note: the dot is used a placeholder here, so you could assume it to be u, or x, or any variable) Assume that we are going to carry out back propagation and a gradient descent step. Suppose that Ne = 0.5, and that x = 0.9, and that the learning rate is equal to 0.01. You can assume that b > 0. What will be the value of wt+1? note: be sure to carry out your answer to at least two (2) decimal places, i.e. 1e-2 precision)

Computer Networking: A Top-Down Approach (7th Edition)
Computer Networking: A Top-Down Approach (7th Edition)
7th Edition
ISBN:9780133594140
Author:James Kurose, Keith Ross
Publisher:James Kurose, Keith Ross
Chapter1: Computer Networks And The Internet
Section: Chapter Questions
Problem R1RQ: What is the difference between a host and an end system? List several different types of end...
See similar textbooks
icon
Related questions
Question

5

Backward Step (i.e. back propagation) and Gradient Descent When training a neural network model we use back propagation to update the weight/bias parameters using gradient descent (or some variation of it). Gradient descent utilizes the chain rule, since each layer of a neural network can be described by a function, and the activation function (and additional multiple layers) can be described by a composition of functions. Suppose that we have the following functions: fi (x, w, b) = w·x + b f2 (•) ReLU (·) f (-) f2 (fi (:)) (note: the dot is used a placeholder here, so you could assume it to be u, or x, or any variable) Assume that we are going to carry out back propagation and a gradient descent step. Suppose that Wt = 0.5, and that x = 0.9, and that the learning rate is equal to 0.01. You can assume that b > 0. What will be the value of we+1? (note: be sure to carry out your answer to at least two (2) decimal places, i.e. 1e-2 precision)
Transcribed Image Text:Backward Step (i.e. back propagation) and Gradient Descent When training a neural network model we use back propagation to update the weight/bias parameters using gradient descent (or some variation of it). Gradient descent utilizes the chain rule, since each layer of a neural network can be described by a function, and the activation function (and additional multiple layers) can be described by a composition of functions. Suppose that we have the following functions: fi (x, w, b) = w·x + b f2 (•) ReLU (·) f (-) f2 (fi (:)) (note: the dot is used a placeholder here, so you could assume it to be u, or x, or any variable) Assume that we are going to carry out back propagation and a gradient descent step. Suppose that Wt = 0.5, and that x = 0.9, and that the learning rate is equal to 0.01. You can assume that b > 0. What will be the value of we+1? (note: be sure to carry out your answer to at least two (2) decimal places, i.e. 1e-2 precision)
Expert Solution
steps

Step by step

Solved in 3 steps with 1 images

SEE SOLUTIONCheck out a sample Q&A here
Blurred answer
Recommended textbooks for you
Computer Networking: A Top-Down Approach (7th Edi…
Computer Networking: A Top-Down Approach (7th Edi…
Computer Engineering
ISBN:
9780133594140
Author:
James Kurose, Keith Ross
Publisher:
PEARSON
Computer Organization and Design MIPS Edition, Fi…
Computer Organization and Design MIPS Edition, Fi…
Computer Engineering
ISBN:
9780124077263
Author:
David A. Patterson, John L. Hennessy
Publisher:
Elsevier Science
Network+ Guide to Networks (MindTap Course List)
Network+ Guide to Networks (MindTap Course List)
Computer Engineering
ISBN:
9781337569330
Author:
Jill West, Tamara Dean, Jean Andrews
Publisher:
Cengage Learning
Concepts of Database Management
Concepts of Database Management
Computer Engineering
ISBN:
9781337093422
Author:
Joy L. Starks, Philip J. Pratt, Mary Z. Last
Publisher:
Cengage Learning
Prelude to Programming
Prelude to Programming
Computer Engineering
ISBN:
9780133750423
Author:
VENIT, Stewart
Publisher:
Pearson Education
Sc Business Data Communications and Networking, T…
Sc Business Data Communications and Networking, T…
Computer Engineering
ISBN:
9781119368830
Author:
FITZGERALD
Publisher:
WILEY
SEE MORE TEXTBOOKS