University of Tunis
Tunis Business School Spring 2025
Tutorial 3
Solutions
Question 1
1. N = 3 steps
Backward Approach
Sn= Available funds before deciding for clinics n, n+1,…,3; n = 1, 2, 3.
Xn = Funds assigned to clinic n; n = 1, 2, 3.
Rn(Sn,Xn) = projected lives saved if Xn million TND are assigned to clinic n; n = 1, 2, 3.
Fn(Sn,Xn) = Maximum total projected saved lives for clinics n, n+1,…,3 given that Xn
million TND are assigned to clinic n from an available budget Sn million TND
before decision; n = 1, 2, 3.
Fn*(Sn) = Fn(Sn,Xn*) = MaxXn Fn(Sn,Xn); n = 1, 2, 3.
Fn(Sn,Xn) = Rn(Sn,Xn) + F*n+1(Sn-Xn); n = 1, 2.
F3(S3,X3) = R3(S3,X3).
Clinic 3
Optimal
F3(S3,X3)
Decision
X3 0 1 2 3 4
F*3(S3) X3*
S3
0 0 0 0
1 10 10 1
2 15 15 2
3 20 20 3
4 30 30 4
Clinic 2
F2(S2,X2) = R2(S2,X2) + F*3(S2-X2) Optimal Decision
X2 0 1 2 3 4
F*2(s2) X 2*
S2
0 0+0=0 0 0
1 0+10=10 15+0=15 15 1
2 0+15=15 15+10=25 20+0=20 25 1
3 0+20=20 15+15=30 20+10=30 24+0=24 30 1 or 2
4 0+30=30 15+20=35 20+15=35 24+10=34 27+0=27 35 1 or 2
Clinic 1
F1(S1,X1) = R1(S1,X1) + F*2(S1-X1) Optimal Decision
X1 0 1 2 3 4
F*1(s1) X1*
S1
4 0+35=35 7+30=37 10+25=35 15+15=30 25+0=25 37 1
Two optimal solutions:
• Allocate 1 million to clinic 1, 1 million to clinic 2, and 2 millions to clinic 3, or
• Allocate 1 million to clinic 1, 2 millions to clinic 2, and 1 million to clinic 3
The total expected number of lives to be saved = 41.
2. The recursive function if a forward approach is adopted is :
Fn*(Sn) = Fn(Sn,Xn*) = MaxXn Fn(Sn,Xn); n = 1, 2, 3.
Fn(Sn,Xn) = Rn(Sn,Xn) + F*n-1(Sn+Xn); n = 2, 3
F1(S1,X1) = R1(S1,X1).
Question 2
Di = requirement of month i with i = 1,2,3,4
Si: Number of spares available at the beginning of the month i for the period i, i+1, i+2, …4,
with i = 1,2,3,4
Xi = Number of spares to buy during the month i with i = 1,2,3,4
Note that any excess in inventory of at least 2 spares will cost more than the fixed setup cost of
500TND. Hence, the only rational inventory levels of spares at the start of each month can only
be 0 or 1.
500 +200Xi + 350Si if Xi 0;
Ri(Si,Xi) = i = 1, 2, 3, 4.
350Si if Xi = 0;
Fi(Si,Xi) = Ri(Si,Xi) + F*i+1(Si+Xi -Di); i = 1, 2, 3, 4.
F4(S4,X4) = R4(S4,X4);
F*i (Si) = MinXi Fi(Si,Xi)
Month 4:
X4
S4 F4 (S4,X4) = R4(S4,X4) Optimal Decision
9 10 F*4(S4) X*4
0 - 2500 2500 10
1 2650 - 2650 9
Month 3:
X3
S3 F3 (S3,X3) Optimal Decision
8 9 10 F*3 X*3
0 - 2300+2500=4800 2500+2650=5150 4800 9
1 2450+2500=4950 2650+2650=5300 - 4950 8
Month 2:
X2 F2 (S2,X2) Optimal Decision
S2
7 8 9 F*2 X*2
0 - 2100+4800=6900 2300+4950=7250 6900 8
1 2250+4800=7050 2450+4950=7400 - 7050 7
Month 1:
X1
S1
F2 (S2,X2) Optimal Decision
10 11 F*1 X*1
0 2500+6900=9400 2700+7050=9750 9400 10
The optimal inventory plan of the next 4 months is to buy exactly what is required. Therefore,
we have to buy 10 spares for the first and last months and respectively, 8 and 9 spares for the
second and third months.
Question 3
N=3
Backward approach
Xi: Number of workers to make available at week i; i = 1, 2, 3.
Si: Number of workers already available at the beginning of week i; i = 1, 2, 3.
Ri(Si, Xi) = The cost of making Xi workers available during week i given that Si workers were
available at the beginning of the week; i = 1, 2, 3.
100xXi + 60(Si – Xi) si Si Xi ;
Ri(Si, Xi) =
100xXi + 40(Xi - Si) si Si Xi ; i = 1, 2,3.
Fi(Si, Xi) = Ri(Si, Xi) + Fi+1*(Xi)
F*i(Si) = MinXi Fi(Si,Xi)
F3(S3, X3) = R3(S3, X3)
Week 3:
F3 (S3,X3) Optimal Decision
X3 7
F3*(S3) X3*
S3
6 100*7+40*1=740 740 7
7 100*7=700 700 7
Week 2:
F2 (S2,X2)= R2(S2, X2) + F3*(X2) Optimal Decision
X2 6 7
F2*(S2) X2*
S2
8 100*6+60*2+740=1460 100*7+60*1+700=1460 1460 6 or 7
Week 1:
F1(S1, x1) Décision optimale
X1 8 F1*(S1) X2*
S1
6 100*8+40*2+1460=2340 2340 8
The company has two solutions: either it recruits 2 workers during the first week, fires 2
workers during the second week, and recruits 1 worker for the 3rd week; or it recruits 2 workers
during the first week, fires 1 worker during the second week, and preserve all the 7 workers for
the 3rd week. The related cost is 2340 TD.
Question 4
N=3
Forward Approach
Sn = The number of components remaining after deciding for systems 1, 2, …. n; n = 1,..,3.
Xn = The number of components integrated in the system n; n = 1,.., 3.
Rn(Sn,Xn) = probability of failure of system n knowing that we have integrated Xn components
into it and Sn components still be available after the decision; n = 1,.., 3.
Fn(Sn,Xn) = probability of failure of the systems 1, …, n; knowing that Sn components remains
available after having integrated Xn components in system n.
Fn(Sn,Xn) = Rn(Sn,Xn) x Fn-1(Sn+Xn)
F1(S1,X1) = R1(S1,X1)
Fn*(Sn) = Fn(Sn,Xn*) = MinXn Fn(Sn,Xn),
Step 1 (System 1)
F1(S1,X1) = R1(S1,X1) Optimal Decision
X1 1 2 3
F1 *(S1) X1*
S1
2 - - 0.04 0.04 3
3 - 0.15 - 0.15 2
4 0.5 - - 0.5 1
Step 2 (System 2/System 1)
F2(S2,X2) = R2(S2,X2) x F1(S2+X2) Optimal
Decision
X2 1 2 3
F2*(S2) X2*
S2
1 0.6x0.04=0.024 0.2x0.15=0.03 0.1x0.5=0.05 0.024 1
2 0.6x0.15=0.09 0.2x0.5=0.10 - 0.09 1
3 0.6x0.5=0.30 - - 0.30 1
Step 3 (System 3/ Systems 1 & 2)
F3(S3,X3) = R3(S3,X3) x F2(S3+X3) Optimal
Decision
X3 1 2 3
F3*(S3) X3*
S3
0 0.4x0.024=0.0096 0.25x0.09=0.0225 0.10x0.30=0.03 0.0096 1
Optimal solution : 3 components for system 1, 1 component for system 2 and 1 component
for system 3.
The reliability of the product = 99.04 % (1-0.0096=0.9904).
Question 5
N=2
Backward approach
Xi = 0 if no project of type i is selected; 1 if the first project of type i is selected; 2 if the second
project of type i is selected for projects of type i, i = A, B
Si is available budget before the decision; i = A, B
Ri(Si,Xi) = Net profit generated by the decision Xi, given that a budget of Si is available before
the decision; i = A, B
F1(S1,X1) = Optimal total net profit from project of both types, given that a budget of S1 is
available before the decision X1.
F1(S1,X1) = R1(S1,X1) + F2*(S1 - C(X1))
where C(X1) is the cost incurred with the decision for project of type 1.
Fi*(Si) = MaxXi Fi(Si, Xi) ; i = 1, 2
F2 (S2,X2) = R2(S2,X2)
Step 2 (B)
F2(S2, X2) = R2(S2,X2) Optimal Decision
X2 0 1 2
F2*(S2) X2*
S2
2 0 1 - 1 1
4 0 1 4 4 2
6 0 1 4 4 2
Step 1 (A)
F1(S1, X1) = R1(S1,X1) + F2(S1-C(X1))
X1 0 1 2
F1*(S1) X1*
S1
6 0+4=4 2+4=6 5+1=6 6 1 or 2
Two possible solutions : Either invest in A1 and B2 ; or invest in A2 and B1 with a total profit
equal to 6.
Question 6
N=3
Given the constraint 4X1+ 2X2+ 3X3 6, we identify the possible values that can be assigned to
each of the three variables. Thus X1{0,1}, X2{0,1,2,3}, X3{0,1,2}
Sn= The rest of the constraint’s RHS before assigning a value to variable Xn
Let’s denote by.
cn the coefficient of Xn in the objective function
an the coefficient of Xn in the constraint
Three terms constitute the objective function. Each term is related to one of the three variables.
Rn(Sn,Xn) = the term from the objective function related to X n given Sn is the rest of the
constraint’s RHS before assigning a value to variable Xn; n = 1, …, 3.
Rn(Sn,Xn) = cnXn
Fn(Sn,Xn) = maximum cumulative terms from the objective function given Sn is the rest of the
constraint’s RHS before assigning a value to variable Xn; n = 1, …, 3.
Fn(Sn,Xn) = Rn(Sn,Xn) + F*n+1(Sn+1) = cnXn + F*n+1(Sn-anXn); n = 1, 2
Where Fn*(Sn) = Fn(Sn,Xn*) = maxXn Fn(Sn,Xn),
F3(S3,X3) = R3(S3,X3)
Step 3 (X3)
F3(S3,X3) = 2X3
X3 0 1 2
F3*(s3) X 3*
S3
0 0 - - 0 0
2 0 - - 0 0
4 0 2 - 2 1
6 0 2 4 4 2
Step 2 (X2)
F2(S2,X2)= 7X2 + F*3(S2-2X2)
X2 0 1 2 3
F*2(S2) X 2*
S2
2 0+0=0 7+0=7 - - 7 1
6 0+4=4 7+2=9 14 + 0 =14 21 + 0 = 21 21 3
Step 1 (X1)
F1(S1,X1)= 3X1 + F*2(S1-4X1)
X1 0 1
F*1(S1) X1*
S1
6 0 + 21 = 21 3 + 7 = 10 21 0
Optimal solution X1 = 0, X2 = 3, X3 = 0. Z*=21
Question 7
N=4
Backward Approach
Sn : Age of the machine at the beginning of year n, S n = 1, 2, 3 ; n = 2,…, 4
The 1st year a new machine is purchased thus S1 = 0.
1 if the machine is replaced at the beginning of year n
Xn = n = 2,…, 4
0 otherwise
X1 = 0, since we start with a new machine.
Rn(Sn,Xn) = the net cost of using the machine during year n.
- SV(Sn) + 10000 + 900 = - SV(Sn) + 10900 if Xn = 1
Rn (Sn,Xn) = n = 2, 3
M(Sn+1) if Xn = 0
Where
SV(Sn): The salvage value of the machine after Sn years of use.
M(Sn): The maintenance cost of the machine of age Sn.
R1 (S1,X1) = 10000 + 900 = 10900.
-SV(S4)+10000+900-SV (1)=-SV(S4)+10900–6000 = -SV(S4)+4900 if X4 = 1
R4 (S4,X4)=
M(S4+1) - SV(S4+1) if X4 = 0
Fn(Sn,Xn) = the minimum total cost of using the machine during years n, n+1, …, 4, knowing
that the age of the available machine at the beginning of year n is S n and for which the decision
Xn is taken.
Fn(Sn,Xn) = Rn (Sn,Xn) + Fn+1*(Sn+1) n = 1, 2, …, 3
1 if Xn = 1
Sn+1= n = 1, 2,…, 3
Sn+1 if Xn = 0
Fn*(Sn) = MinXn Fn(Sn,Xn)
F4(S4,X4) = R4(S4,X4)
Step 4 (Year 4) F4(S4,X4) = R4(S4,X4)
- SV(S4) + 4900 if X4 = 1
R4 (S4,X4) =
M(S4+1) - SV (S4+1) if X4 = 0
F4 (S4,X4) Optimal Decision
X4 1 0
F4*(S4) X4*
S4
1 -6000 + 4900 = - 1100 1200 -3000 = - 1800 -1800 0
2 -3000 + 4900 = 1900 2500 -500 = 2000 1900 1
3 -500 + 4900= 4400 - 4400 1
Step 3 (Year 3 / Year 4)
-SV(S3) +10900+F4*(1) = -SV(S3) +10900 -1800 = -SV(S3) + 9100 if X3 = 1
F3 (S3,X3) =
M(S3+1) + F4*(S3+1) if X3 = 0
F3 (S3,X3) Optimal Decision
X3 1 0
F3* (S3) X3*
S3
1 -6000 + 9100 = 3100 1200 + 1900 = 3100 3100 1 or 0
2 -3000 + 9100 = 6100 2500 + 4400 = 6900 6100 1
Step 2 (Year 2 / Years 3 and 4)
-VR(S2) +10900 +F3*(1) = -VR(S2) + 10900 + 3100 = -VR(S2) + 14000 if X2 = 1
F2 (S2,X2)=
M(S2+1) + F3*(S2+1) if X2 = 0
F2 (S2,X2) Optimal Decision
X2 1 0
F2(S2) X2*
S2
1 -6000 + 14000 = 8000 1200 + 6100 = 7300 7300 0
Step 1 (Year 1 / Years 2, 3 and 4)
No decision is needed to be taken during the first year.
F1*(0) = F1(0,0) = 10900 + F2*(1) = 10900 + 7300 = 18100
Optimal planning: The purchased machine at the beginning of the 1st year should be replaced
at the beginning of the 3rd year and then sold at the end of the 4th year.
Total cost = 18100 dinars.
