(a) Suggest a network architecture appropriate for MNIST and explain why the cross- entropy cost function...
Fantastic news! We've Found the answer you've been seeking!
Question:
Transcribed Image Text:
(a) Suggest a network architecture appropriate for MNIST and explain why the cross- entropy cost function might be preferable to the mean-squared-error when the sigmoid activation function is used. (5 marks) (b) Explain the problem of overtraining a neural network and how it could be alleviated. How does MNIST address this problem? (5 marks) (c) Briefly explain what gradient descent means and how it relates to the cost or loss function. Explain what each of the following symbols represents and the relationship between then in gradient descent. w, and ac aw Vij ? (d) At the core of the backpropagation algorithm are the matrix formulae: ac = 8} ak awk 1-1 s = (a - y) o' (z) 8 = ((w+1) 8+1) o' (2) Briefly explain what each term means and which one in particular relates to backpropagation. (5 marks) Also, for a three layer network, write a Python code fragment which includes each of these formulae. (10 marks) (a) Suggest a network architecture appropriate for MNIST and explain why the cross- entropy cost function might be preferable to the mean-squared-error when the sigmoid activation function is used. (5 marks) (b) Explain the problem of overtraining a neural network and how it could be alleviated. How does MNIST address this problem? (5 marks) (c) Briefly explain what gradient descent means and how it relates to the cost or loss function. Explain what each of the following symbols represents and the relationship between then in gradient descent. w, and ac aw Vij ? (d) At the core of the backpropagation algorithm are the matrix formulae: ac = 8} ak awk 1-1 s = (a - y) o' (z) 8 = ((w+1) 8+1) o' (2) Briefly explain what each term means and which one in particular relates to backpropagation. (5 marks) Also, for a three layer network, write a Python code fragment which includes each of these formulae. (10 marks)
Expert Answer:
Answer rating: 100% (QA)
a For MNIST a straightforward and successful organization design could be a feedforward brain networ... View the full answer
Related Book For
Posted Date:
Students also viewed these programming questions
-
Free vibration of two rotor system The flywheel of an engine driving a dynamo has mass of 200kg and has a radius of gyration of 30cm. The shaft at the flywheel end has as an effective length of 25cm...
-
QUIZ... Let D be a poset and let f : D D be a monotone function. (i) Give the definition of the least pre-fixed point, fix (f), of f. Show that fix (f) is a fixed point of f. [5 marks] (ii) Show that...
-
Let A, B be sets. Define: (a) the Cartesian product (A B) (b) the set of relations R between A and B (c) the identity relation A on the set A [3 marks] Suppose S, T are relations between A and B, and...
-
What is the difference between an optimistic approach and a pessimistic approach to decision making under assumed uncertainty?
-
Let f be the matrix transformation defined in Example 5. Find and sketch the image of the rectangle with vertices (0, 0), (1, 0), (1, l), and (0, 1) for h = 2 and k = 3.
-
Code a constructor that get the full name, id number, cell number and gender via parameters. The constructor should initialise the points and number of purchases to zero.
-
On December 31, 2017, Kim Company issued \(\$ 500,000\) of five-year, 12 percent bonds payable for \(\$ 538,609\), yielding an effective interest rate of ten percent. Interest is payable semiannually...
-
On December 31, Year 1, Kelly Corporation of Toronto paid 13.7 million Libyan dinars (LD) for 100% of the outstanding common shares of Arkenu Company of Libya. On this date, the fair values of...
-
A sampling distribution is a) a relative frequency distribution b) a probability distribution c) both a and b d) neither of the above
-
International Technology Inc. (ITI) acquired all of the voting stock of Global Outsourcing Corporation (GOC) on June 30,2010, for $ 110 million in cash and stock, plus an earnings contingency payable...
-
Problem 6.1. The probability distribution of returns on stocks X and Y follows. Probabilit 0.2 0.3 0.3 0.2 y Rx -20 0% 20 40 % % % Ry 1% 3% 5% 7% Calculate the expected return and the standard...
-
Is there any circumstance where a less effective new treatment might be adopted? Explain.
-
1. Write a C program which can create a phone structure that consists of screen size, RAM and price of a phone. In the main function, input some phones' information using that structure and display...
-
There are four key components of a company's financial statements: Income Statement, Statement of Equity/Capital (Changes in Capital/Equity), Balance Sheet (Statement of Financial Position), and...
-
Macy Inc. produces the following three items of stuffed toys using an automated technology: Bear, Puppy, and Bunny. The estimation for the number of units produced, machine hours used, and direct...
-
What are the psychosocial support model under social care, specialist care, and supportive care?
-
A film director and a scientist want to study which elements make a movie scary. They show participants various film clips and measure "scariness" by changes in participants' heart rates or in how...
-
What can you do to reduce hunger where you live? To reduce hunger globally?
-
The complex gas-phase reactions are elementary and carried out in a PFR with a heat exchanger. Pure A enters at a rate of 5 mol/min, a concentration of 0.2 mol/dm 3 , and temperature 300 K. The...
-
The enzymatic hydrolysis of starch was carried out with and without maltose and -dextrin added. (Adapted from S. Aiba, A. E. Humphrey, and N.F. Mills, Biochemical Engineering, New York: Academic...
-
How do the steps in the design of a CSTR differ from those of a CSTR or a PFR with pressure drop?
-
Part 1: Calculating Short-Term Measurements The following annual information is available for Lakewood Industries, an investment center: a. Calculate ROI (return on investment). Base your...
-
Four years ago, based on a pre-tax NPV analysis, Harper Inc. decided to add new equipment with a cost of \(\$ 85,000\), allowing the company to expand its product offerings. The data used in the...
-
For each of the following performance measures, indicate (a) whether the measure is an internal performance measure or an external performance measure and (b) whether the measure primarily relates to...
Study smarter with the SolutionInn App