Soft Computing Techniques
Week-07 Assignment-07 (Solution)
1. (b). In a multi-objective optimization problem, a solution is said to ‘dominate’ another solution if it
is better in atleast one objective and worse in none.
2. (a). Let Xt be the travel time and Xp be the price of food of a restaurant X. We need to minimize
both the travel time and the price of food. Now, from the given table we have,
Ct < Bt and Cp < Bp
So, C dominates B. Therefore, (a) is correct and (b) is wrong. For option (c) we have,
Dt < At and Dp > Ap
So, we cannot conclude anything about the domination of D over A. Therefore, (c) is also wrong.
3. (a). The non-dominated set is a set of all solutions that are not dominated by any member of the
solution set.
4. (d). Stagnation occurs when the algorithm stop making significant progress toward better solutions
due to poor population quality.
5. (b). For the given data, using weighted average method the optimizing function becomes,
f (x) = w1 f1 (x) + w2 f2 (x)
=⇒ f (x) = 0.7(x − 2)2 + 0.3(x2 − 3)
=⇒ f (x) = x2 − 2.8x + 1.9
By differentiating f (x) [Link] x and equating it to zero then we get,
f ′ (x) = 0
d 2
=⇒ x − 2.8x + 1.9 = 0
dx
=⇒ 2x − 2.8 = 0
=⇒ x = 1.4
6. (c). The function f (x) has a minima at x = a if f ′′ (a) > 0.
7. (c). The optimal point we have found in question 5 is x = 1.4, so the optimal values of the function
f1 (x) and f2 (x) are,
f1 (1.4) = (1.4 − 2)2
=⇒ f1 (1.4) = 0.36
and f2 (1.4) = 1.42 − 3
=⇒ f2 (1.4) = −1.04
8. (a). A large population size ensure greater genetic diversity, reducing the risk of premature convergence.
9. (d). For the given functions, the optimal points of f1 (x) and f2 (x) are given by,
d
f1 (x) = 0
dx
d
(x − 2)2 = 0
=⇒
dx
=⇒ 2(x − 2) = 0
=⇒ x = 2 ∈ [0, 4]
d
and f2 (x) = 0
dx
d 2
=⇒ x −3 =0
dx
=⇒ 2x − 0 = 0
=⇒ x = 0 ∈ [0, 4]
Now, we have f1′′ (2) > 0 and f2′′ (0) > 0, so f1 (x) has a minimum at the point x = 2 and f2 (x) has a
minimum at the point x = 0. Therefore the ideal solution is given by,
z1∗ = (2 − 2)2
=⇒ z1∗ = 0
and z2∗ = 02 − 3
=⇒ z2∗ = −3
So, we get, z∗ = (0, −3)
10. (a). Now, we have to find the nadir solution i.e., the worst point. Since the problem is to minimize the
functions so the worst point will be the point where the given functions attains their maximum in the
feasible region [0, 4]. We know that the functions are increasing functions so the maximum will occur
at x = 4.
f1 (4) = (4 − 2)2
=⇒ f1 (4) = 4
f2 (4) = 42 − 3
=⇒ f1 (4) = 13
=⇒ zN = (4, 13)
11. (d). The Global criterion function G(x) is given by,
s
f (x) − z ∗ 2 f (x) − z ∗ 2
1 1 2 2
G(x) = +
z1N − z1∗ z2N − z2∗
s
((x − 2)2 − 0)2 (x2 − 3 + 3)2
= +
(4 − 0)2 (13 + 3)2
r
(x − 2)4 x4
=⇒ G(x) = +
16 256
12. (a). Truncation selection is based on assigning a selection probability proportional to the fitness of an
individual.
13. (b). The primary advantage of rank-based selection over fitness-proportionate selection is that it avoids
the dominance of highly fit individuals.
14. (b). In Weighted sum method, it is difficult to find certain pareto optimal solutions in case of a
non-convex objective space.
