CS542 – Lab7
Homework 4 solutions and practice midterm
Homework 4 solutions
CS542_HW4_solution.ipynb
Exam sub-topics
Assume we have “n” buckets and many balls. We drop the balls randomly in buckets. How many balls should we drop to fill all the buckets in expectation? (Assume n >> 1)
Maximize “|| X ||_1” given that “|| X ||_2 = 1”.
What is the optimal learning rate for gradient descent in linear regression?
Let’s say we have D different classifiers {f_1, …, f_D}. Based on a training data {(X_1,y_1), …, (X_n, y_n) we choose a classifier f^* which has the minimum classification error = 𝛂* . What is 95% interval for the test error of f^*?
What is an unbiased estimator for 𝛃 in Q5.1 in the midterm practice exam?
Practice midterm solutions
CS542_Fall2020_Midterm_Practice_with_Solutions.pdf
Any questions concerns?