23/03/2024, 14:53 Evolutionary Computation For Single And Multi-Objective Optimization - - Unit 11 - Week 8 : EC Techniques for Multi-Objecti…

(https://swayam.gov.in) (https://swayam.gov.in/nc_details/NPTEL)

[email protected] 

NPTEL (https://swayam.gov.in/explorer?ncCode=NPTEL) » Evolutionary Computation For Single And Multi-Objective Optimization (course)

If already registered,
click to check your
payment status
Assignment 8
The due date for submitting this assignment has passed.
Due on 2024-03-20, 23:59 IST.
As per our records you have not submitted this assignment.
Course outline
1) The main features of SPEA2 are the following 1 point
About NPTEL ()
(a) a fine-grained fitness assignment strategy

How does an (b) an enhanced archive truncation method

NPTEL online (c) non-dominated sorting
course work? () (d) crowding distance

Week 0 : No, the answer is incorrect.

Score: 0
Prerequisite ()
Accepted Answers:
(a) a fine-grained fitness assignment strategy
Week 1: (b) an enhanced archive truncation method
Introduction and
Principles of
2) Considering M as the number of objectives and Q = N + N ¯ ¯
, where N is the population size, and N is the 1 point
Evolutionary
Computation (EC) archive size, the computational complexity of the fine-grained fitness assignment strategy of SPEA2 is
()
(a) O(M Q2 )
Week 2 : Binary-
Coded Genetic (b) O(M Q logQ)
Algorithm ()
(c) O(Q2 )
Week 3 : Real-
Coded Genetic (d) O(Q2 logQ)

Algorithm ()
No, the answer is incorrect.
Score: 0
Week 4: Other Accepted Answers:
Evolutionary (c) O(Q2 )
Computation
Techniques () 3) The raw fitness of SPEA2 is calculated as R(i) = ∑ S(j) for solution (i) . Here, S(j) is the strength 2 points
j∈Q(t), i≻j

Week 5: Constraint of solution (j) that is calculated as S(j) = |{i|i ∈ Q(t) ∧ i ≻ j}| Q(t) . is the current population and the symbol (i ≻ j)
Handling with EC represents solution (j) dominates solution (i) . The objective function values of a bi-objective optimization problem in the
Techniques ()
current population are given as
(1) T (2) T (3) T (4) T (5)
f = (0.139, 0.661) , f = (0.286, 0.378) , f = (0.500, 0.418) , f = (0.632, 0.495) , f = (0.740, 0.725)
Week 6:
f
(6)
= (0.911, 0.195)
T
. Here, both the objectives are of minimization type. The raw fitness of solution 4 (f
(4)
) is
Introduction to
Multi-Objective (a) 0
Optimization ()
(b) 3

Week 7: Classical (c) 5

and EC Multi- (d) 7
Objective

N h i i

https://onlinecourses.nptel.ac.in/noc24_me30/unit?unit=60&assessment=109 1/3
23/03/2024, 14:53 Evolutionary Computation For Single And Multi-Objective Optimization - - Unit 11 - Week 8 : EC Techniques for Multi-Objecti…

Optimization No, the answer is incorrect.

Techniques () Score: 0
Accepted Answers:
(c) 5
Week 8 : EC
Techniques for 4) Consider the objective function values of a bi-objective optimization of six solutions, 1 point
Multi-Objective f
(1)
= (0.578, 4.387)
T
, f
(2)
= (0.275, 1.998)
T
, f
(3)
= (0.905, 0.299)
T
, f
(4)
= (1.413, 5.075)
T
, f
(5)
= (1.546, 1.466)
Optimization () and f (6) T
. Here, both the objectives are of minimization types. Using k− th nearest neighbor method of
= (1.207, 3.337)

SPEA2, what is the density of solution 3 (f

(3)
= (0.905, 0.299)
T
) ? Assume k = 4 .
Lec 23 : Strength
Pareto Evolutionary
(a) 0.179
Algorithm: SPEA2:
Introduction (unit? (b) 0.346
unit=60&lesson=61) (c) 0.615
Lec 24 : Strength (d) 0.956
Pareto Evolutionary
No, the answer is incorrect.
Algorithm: SPEA2: Score: 0
Simulations (unit?
Accepted Answers:
unit=60&lesson=62)
(b) 0.346
Lec 25 : Performance
Assessment of Multi- 5) In SPEA2, when the number of solutions (having fitness (F (i) < 1) ) is more than the size of the archive, the 1 point
Objective EC archive is filled using
Techniques (unit?
unit=60&lesson=63)
(a) solutions having fitness (F (i)) < 1 and then copy solutions having fitness (F (i)) > 1 in their ascending order of
Lec 26 : Closure of fitness values till the archive is filled
EC for Single and
Multi-Objective
(b) solutions having fitness (F (i)) < 1 and then copy solutions having fitness (F (i)) > 1 till the archive is filled
Optimization (unit?
unit=60&lesson=64)
(c) solutions having fitness (F (i)) > 1 in their ascending order of fitness values till the archive is filled
Quiz: Assignment 8
(assessment?
(d) truncation error
name=109)
No, the answer is incorrect.
Feedback Form08 Score: 0
(unit? Accepted Answers:
unit=60&lesson=65)
(d) truncation error
Solution : Assignment
8 (unit?
unit=60&lesson=66) 6) The performance assessment of EC techniques for multi-objective optimization is needed because 1 point

Text transcripts () (a) EC techniques are stochastic in nature and can generate different set of non-dominated solutions in different runs
(b) Different EC techniques can be compared on the given set of problems
Books () (c) EC techniques are developed using ideal-multi-objective optimization approach
(d) One solution can be selected from the obtained set of non-dominated solutions from EC techniques

No, the answer is incorrect.

Score: 0
Accepted Answers:
(a) EC techniques are stochastic in nature and can generate different set of non-dominated solutions in different runs
(b) Different EC techniques can be compared on the given set of problems

7) Hypervolume measures 1 point

(a) that portion of the objective space that can be maximized.

(b) that portion of the objective space that dominates an approximate set A, and is to be maximized.

(c) that portion of the objective space that is weakly dominated by an approximate set A, and is to be minimized.

(d) that portion of the objective space that is weakly dominated by an approximate set A, and is to be maximized.

No, the answer is incorrect.

Score: 0
Accepted Answers:
(d) that portion of the objective space that is weakly dominated by an approximate set A, and is to be maximized.

8) 2 points

https://onlinecourses.nptel.ac.in/noc24_me30/unit?unit=60&assessment=109 2/3
23/03/2024, 14:53 Evolutionary Computation For Single And Multi-Objective Optimization - - Unit 11 - Week 8 : EC Techniques for Multi-Objecti…

Consider the objective function values of a bi-objective optimization of five solutions,

f
(1)
= (4.954, 0.786)
T
, f
(2)
= (2.568, 2.351)
T
, f
(3)
= (6.467, 0.330)
T
, f
(4)
= (3.698, 1.526)
T
, and
f
(5)
= (2.929, 2.012)
T
. Here, both the objectives are of minimization types. Hypervolume of these solutions with respect
T
to W = (1.1, 1.1) is lying between

(a) 0 to 0.25
(b) 0.26 to 0.50
(c) 0.51 to 0.75
(d) 0.76 to 1.0
No, the answer is incorrect.
Score: 0
Accepted Answers:
(c) 0.51 to 0.75

9) Using hypervolume indicator for assessment of the given bi-objective optimization problem, set A is better than 1 point
set B in

(a) extent
(b) proximity
(c) diversity
(d) evenness

No, the answer is incorrect.

Score: 0
Accepted Answers:
(a) extent

10) Consider the objective function values of a bi-objective optimization of five solutions, 1 point
f
(1)
= (0.470, 0.824)
T
, f
(2)
= (0.995, 0.280)
T
, f
(3)
= (2.021, 0.113)
T
, f
(4)
= (0.718, 0.530)
T
, and
= (1.943, 0.138) . Here, both the objectives are of minimization types. Using the truncation operator of SPEA2,
(5) T
f

which solution will be removed from the current set of solutions? Assume k = 3.

(a) solution 1
(b) solution 2
(c) solution 3
(d) solution 4
(e) solution 5

No, the answer is incorrect.

Score: 0
Accepted Answers:
(b) solution 2

https://onlinecourses.nptel.ac.in/noc24_me30/unit?unit=60&assessment=109 3/3