Which of the following is a recommended way to initialize u; in k-means? Set μ E R to a random value between -e and e, for some small €. a. b. c. Od. Pick k distinct random integers i₁, ..., ik from {1,..., m}. Set μ₁ = x(i₁), µ₂ = x(¹₂), ..., µx = x(ix). Mk Pick k distinct random integers i₁, ..., ik from {1,..., k}. Set μ₁ = x(¹₁), μ₂ = x(¹₂),..., Mk = x(ik). Pick a random integer i from {1,..., k}. Set μ₁ = M₂ = ... = Mk = x(i).

Introductory Circuit Analysis (13th Edition)
13th Edition
ISBN:9780133923605
Author:Robert L. Boylestad
Publisher:Robert L. Boylestad
Chapter1: Introduction
Section: Chapter Questions
Problem 1P: Visit your local library (at school or home) and describe the extent to which it provides literature...
icon
Related questions
Question

help for the 2 mcq

Which of the following is a recommended way to initialize µ; in k-means?
Set μ¿ E R¹ to a random value between-e and e, for some small €.
a.
b.
C.
d.
·9 ... 9
Pick k distinct random integers i₁, ..., ik from {1,
Set μ₁ = x(¹₁), µ₂ = x(²₂),..., Mk = x(ik).
m}.
Pick k distinct random integers i₁, ..., ik from {1,
Set μ₁ = x(¹₁), µ₂ = x(¹₂),..., µk = .
=
Mk x(
x(ik).
Pick a random integer i from {1,..., k}. Set μ₁ = M₂ = = µk = x(¹),
k}.
Transcribed Image Text:Which of the following is a recommended way to initialize µ; in k-means? Set μ¿ E R¹ to a random value between-e and e, for some small €. a. b. C. d. ·9 ... 9 Pick k distinct random integers i₁, ..., ik from {1, Set μ₁ = x(¹₁), µ₂ = x(²₂),..., Mk = x(ik). m}. Pick k distinct random integers i₁, ..., ik from {1, Set μ₁ = x(¹₁), µ₂ = x(¹₂),..., µk = . = Mk x( x(ik). Pick a random integer i from {1,..., k}. Set μ₁ = M₂ = = µk = x(¹), k}.
Suppose we have run k-means with k = 5 and k = 10 We find that the cost function value is much larger for k= 10
than k 5. What is the reason?
a There must be a bug in the code.
b. In the run with k = 10 k-means got stuck in a bad local minimum. We should consider re-running k-means
with multiple random initializations.
c. The ideal number of clusters is k = 5
d. In the run with k = 5 k-means got lucky. We should consider re-running k-means
with k 5 and different random initializations until it performs worse than with k = 10
Transcribed Image Text:Suppose we have run k-means with k = 5 and k = 10 We find that the cost function value is much larger for k= 10 than k 5. What is the reason? a There must be a bug in the code. b. In the run with k = 10 k-means got stuck in a bad local minimum. We should consider re-running k-means with multiple random initializations. c. The ideal number of clusters is k = 5 d. In the run with k = 5 k-means got lucky. We should consider re-running k-means with k 5 and different random initializations until it performs worse than with k = 10
Expert Solution
steps

Step by step

Solved in 3 steps

Blurred answer
Recommended textbooks for you
Introductory Circuit Analysis (13th Edition)
Introductory Circuit Analysis (13th Edition)
Electrical Engineering
ISBN:
9780133923605
Author:
Robert L. Boylestad
Publisher:
PEARSON
Delmar's Standard Textbook Of Electricity
Delmar's Standard Textbook Of Electricity
Electrical Engineering
ISBN:
9781337900348
Author:
Stephen L. Herman
Publisher:
Cengage Learning
Programmable Logic Controllers
Programmable Logic Controllers
Electrical Engineering
ISBN:
9780073373843
Author:
Frank D. Petruzella
Publisher:
McGraw-Hill Education
Fundamentals of Electric Circuits
Fundamentals of Electric Circuits
Electrical Engineering
ISBN:
9780078028229
Author:
Charles K Alexander, Matthew Sadiku
Publisher:
McGraw-Hill Education
Electric Circuits. (11th Edition)
Electric Circuits. (11th Edition)
Electrical Engineering
ISBN:
9780134746968
Author:
James W. Nilsson, Susan Riedel
Publisher:
PEARSON
Engineering Electromagnetics
Engineering Electromagnetics
Electrical Engineering
ISBN:
9780078028151
Author:
Hayt, William H. (william Hart), Jr, BUCK, John A.
Publisher:
Mcgraw-hill Education,