MPE Mathematical Problems in Engineering 1563-5147 1024-123X Hindawi 10.1155/2017/5791796 5791796 Research Article A Single-Machine Scheduling Problem with Uncertainty in Processing Times and Outsourcing Costs http://orcid.org/0000-0002-0663-3445 Park Myoung-Ju 1 http://orcid.org/0000-0002-7479-6473 Choi Byung-Cheon 2 Arik Sabri 1 Department of Industrial and Management Systems Engineering Kyung Hee University 1732 Deogyeong-daero Giheung-gu Yongin-si Kyunggi-do 17104 Republic of Korea khu.ac.kr 2 Department of Business Administration Chungnam National University 99 Daehak-ro Yuseong-gu Daejeon 34134 Republic of Korea cnu.ac.kr 2017 532017 2017 13 12 2016 22 02 2017 532017 2017 Copyright © 2017 Myoung-Ju Park and Byung-Cheon Choi. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We consider a single-machine scheduling problem with an outsourcing option in an environment where the processing time and outsourcing cost are uncertain. The performance measure is the total cost of processing some jobs in-house and outsourcing the rest. The cost of processing in-house jobs is measured as the total weighted completion time, which can be considered the operating cost. The uncertainty is described through either an interval or a discrete scenario. The objective is to minimize the maximum deviation from the optimal cost of each scenario. Since the deterministic version is known to be NP-hard, we focus on two special cases, one in which all jobs have identical weights and the other in which all jobs have identical processing times. We analyze the computational complexity of each case and present the conditions that make them polynomially solvable.

Chungnam National University
1. Introduction

We consider a single-machine scheduling problem with an outsourcing option such that the processing time and outsourcing cost are not known in advance. The uncertainty of two parameters is motivated from unpredictable events, such as the introduction of new machines and disruptions in the finished product’s delivery.

In this paper, their uncertainty is described as two types of the scenario set, that is, an interval scenario and a discrete scenario. Let S be the set of scenarios. Let pjs and ojs denote the processing time and the outsourcing cost of job j under scenario sS, respectively. In the interval scenario, the processing time and outsourcing cost of job j are given as any value within the intervals [pjL,pjU] and [ojL,ojU], where pjLmin{pjssS}, pjUmax{pjssS}, ojLmin{ojssS}, and ojUmax{ojssS}. A discrete scenario sS is described as a vector of processing times and outsourcing costs (p1s,,pns;o1s,,ons).

The scheduling problems under the scenario-based uncertainty above have been studied since Daniels and Kouvelis  and Kouvelis and Yu  (see Aissi et al.  for the comprehensive survey). Furthermore, the deterministic version of the single-machine scheduling problem with an outsourcing option has been extensively studied since Vickson  (see Shabtay et al.  for comprehensive survey). To our best knowledge, the scheduling problem with an outsourcing option under uncertainty for some parameters has only been studied by Choi and Chung . The authors considered the case in which the processing time is uncertain while the outsourcing cost is known. They analyzed the computational complexity for various cases, depending on the following:

Whether the performance measure of in-house jobs is the makespan or the total completion time.

Whether the uncertainty is described as the interval scenario or the discrete scenario.

They showed that, for the problem with the makespan as the performance measure of in-house jobs, an interval scenario case is polynomially solvable, while a discrete scenario case is NP-hard. Furthermore, they developed the 2-approximation algorithm based on the linear programming and the fully polynomial-time approximation scheme for the discrete scenario case. For the problem with the total completion time as the performance measure of in-house jobs, referred to as Problem TCO, it is known that both scenario cases are NP-hard. Thus, they focused on the case with a special structure of processing times, pjs=pj+αs, and proved the polynomiality for both scenario cases. Our problem can be considered a general version of Problem TCO, in that the performance of in-house jobs is the total weighted completion time and the outsourcing cost is uncertain.

The formal definition of our problem is as follows: Consider the set J={1,2,,n} of n independent jobs that can be either processed on a single in-house machine or outsourced to a subcontractor. Let wj be the weight of job j. Let (I,σ) be the schedule such that

I and JI are the sets of in-house and outsourced jobs, respectively,

σ is a sequence of jobs in I.

In this paper, the performance measure for in-house jobs is expressed as the total weighted completion time. Let Cjs(σ) be the completion time of an in-house job j in σ under scenario s. Our problem is to find an optimal schedule (I,σ) to minimize (1)ZI,σ=maxsSzsI,σ-zsIs,σs,where

zs(I,σ)=jIwjCjs(σ)+jJIojs;

(Is,σs) is the optimal schedule to minimize zs(I,σ) under scenario s.

Let our problem be referred to as Problem P. For both scenarios, Problem P is NP-hard because the problem of finding (Is,σs) is known to be NP-hard [7, 8]. For the interval scenario case, however, the optimal schedule under a midpoint scenario θ is known to have an approximation factor of two in , where pjθ=(pjL+pjU)/2 and ojθ=(ojL+ojU)/2 for each jJ. Thus, we focus on two cases below:

The weight of each job is identical; that is, wj=1 for each jJ.

The processing time of each job is identical under each scenario; that is, pjs=ps for each jJ and each sS.

Let cases (i) and (ii) be referred to as Problem P1 and Problem P2, respectively. Note that, for two cases, (Is,σs) can be obtained in O(n2) by the algorithm in [7, 8].

The remainder of the paper is organized as follows. Sections 2 and 3 analyze the computational complexity and the polynomially solvable case for Problems P1 and P2. Finally, Section 4 draws conclusions and suggests future research directions.

2. Problem P1

Both scenario cases of Problem P1 without an outsourcing option, that is, the min–max regret version of 1Cj, are NP-hard [2, 12], which implies that both scenario cases of Problem P1 are at least NP-hard, even if the outsourcing costs are certain. In this section, firstly, we prove the NP-hardness of the discrete scenario case with certain processing times. Note that the complexity of the interval scenario case with certain processing times is open. Then, we introduce two polynomially solvable cases.

2.1. NP-Hardness

In this subsection, we prove the NP-hardness for the discrete scenario case with certain processing times. In this case, the sequence of in-house jobs is identical for an optimal schedule under each sS. Thus, throughout the remainder of the section, we use I instead of (I,σ) for notational simplicity.

Theorem 1.

The discrete scenario case of Problem P1 is NP-hard, even if the processing times are certain; that is, pjs=pj for each sS.

Proof.

We reduce the partition problem known to be NP-complete , defined below, to Problem P1: given q positive integers {a1,a2,,aq} such that j=1qaj=A, is there a subset Q such that jQaj=A/2?

Given an instance of the partition problem, we can construct an instance of Problem P1 with two scenarios, as follows: set n=2q and J={1,2,,2q}. For j=1,2,,q,

p2j-1=p2j=Mj,

o2j-11=0 and o2j1=aj,

o2j-12=(4(q-j)+3)Mj and o2j2=(4(q-j)+3)Mj-aj,

where M>0 is a sufficiently large value. Clearly, this reduction can be done in polynomial time. Note the following:

The optimal schedule of scenario 1 is to outsource all jobs. Thus, (2)z1I1=A.

The optimal schedule of scenario 2 is to process all jobs in-house. Thus, (3)z2I2=j=1q4q-j+3Mj.

Henceforth, we show that a schedule I exists such that (4)ZIj=1qq-j+1Mj-A2if and only if there exists a solution to the partition problem.

Suppose that there exists a solution Q- to the partition problem. We can construct a schedule I-={2j-1jQ-}{2jjQ-}. Then, in I-, we have the following:

Exactly one of the jobs in {2j-1,2j},j=1,2,,q, is processed in-house. Thus, under scenarios 1 and 2, the total completion time of in-house jobs is (5)j=1qq-j+1Mj.

The total outsourcing cost is, under scenario 1, (6)jJI-oj1=jQ-aj=A2and, under scenario 2, it is (7)jJI-oj2=j=1q4q-j+3Mj-jQ-aj=j=1q4q-j+3Mj-A2.

Then, (8)ZI-=maxz1I--z1I1,z2I--z2I2=j=1qq-j+1Mj-A2.

Suppose a schedule I^ exists such that (9)ZI^j=1qq-j+1Mj-A2.

Claim. In I^, exactly one of the jobs in {2j-1,2j},j=1,2,,q, is processed in-house.

Proof. Let v be the largest index such that both jobs in {2v-1,2v} are processed in-house and let w be the largest index such that both jobs in {2w-1,2w} are outsourced. If v or w does not exist, then let v= and w= for consistency of notation. Consider the following two cases.

v>w.

In this case, exactly one of the jobs in {2j-1,2j},j=v+1,v+2,,q, is processed in-house. Then, (10)z1I^2q-v+3Mv+j=v+1qq-j+1Mj=q-v+2Mv+j=vqq-j+1Mj.Note that the value in the right-hand side is the lower bound of the total completion time of in-house jobs in I^. Since (q-v+2)>1 and Mv>j=1v-1(q-j+1)Mj+A, however, (11)z1I^-z1I1>j=1qq-j+1Mj>j=1qq-j+1Mj-A2.This is a contradiction.

v<w.

In this case, exactly one of the jobs in {2j-1,2j},j=w+1,w+2,,q, is outsourced. Then, (12)z2I^24q-w+3Mw-aw+j=w+1q4q-j+3Mj-aj+j=w+1qq-j+1Mj=4q-w+3Mw+j=wq4q-j+3Mj-aj+j=w+1qq-j+1Mj.Note that the value in the right-hand side means the total cost of jobs in {2w-1,2w,,2q}I^. Since (13)Mw>j=1w-14q-j+3Mj+j=1w-1q-j+1Mj+A,4q-w+3Mw>q-w+1Mw+Mw,however, (14)z2I^-z2I2>j=1qq-j+1Mj>j=1qq-j+1Mj-A2.This is a contradiction. By cases (i) and (ii), the proof is complete.

Let Q^={j2jI^,j=1,2,,q}. Then, by the above claim, we have (15)z1I^=j=1qq-j+1Mj+jQ^aj,z2I^=j=1qq-j+1Mj+j=1q4q-j+3Mj-jQ^aj.Hence, (16)ZI^=j=1qq-j+1Mj+maxjQ^aj-A,-jQ^aj.This implies that jQ^aj=A/2 and Q^ is the solution to the partition problem.

2.2. Polynomiality

In this subsection, we introduce the conditions that make Problem P1 polynomially solvable. Consider two cases below.

pjs=pj+αs and ojs=oj+βs. In the interval scenario case, the uncertainty values of αs and βs have any values from [αL,αU] and [βL,βU], where αL and αU (βL and βU) are the lower and upper bounds of αs (βs), respectively. On the other hand, the discrete scenario case has the finite set of scenarios S={(α1,β1),(α2,β2),,(αk,βk)}.

pjs=p, and the ordered conditions are satisfied. In the interval scenario case, the ordered conditions can be stated as follows: (17)ojLoj+1L,ojUoj+1Uforj=1,2,,n-1.In the discrete scenario case, for each sS, (18)ojsoj+1sforj=1,2,,n-1.

Note that in-house jobs are processed by increasing order of pj in an optimal schedule for both cases due to the special structure of the processing time uncertainty. Thus, we use I and Cjs instead of (I,σ) and Cjs(σ), respectively, throughout the remainder of the section.

First, we consider case (i). Using the following lemma, the interval scenario case is reduced to the discrete scenario case.

Lemma 2.

For the interval scenario case of Problem P1 with pjs=pj+αs and ojs=oj+βs, Z(I) is determined by a scenario in (19)αL,βL,αL,βU,αU,βL,αU,βU.

Proof.

z s ( I ) can be expressed as follows: (20)zsI=i=1nxij=1ipjxj+αsi=1nxij=1ixj+j=1noj+βs1-xj,where xj=1 if jI and xj=0 otherwise. Thus, zs(I) is a linear function of αs and βs. Since zs(Is) is a minimum function of linear functions, it is a concave function of αs and βs. Thus, zs(I)-zs(Is) is a convex function of αs and βs. Since Z(I) is the maximum of convex functions on S, it is obtained for one scenario in {(αL,βL),(αL,βU),(αU,βL),(αU,βU)}.

Now, we focus on the discrete scenario case. Without loss of generality, we assume that p1p2pn. Let I0(g) be the schedule that minimizes z0(I) among schedules with g in-house jobs, where z0(I) is the total cost of I under (αs,βs)=(0,0). Let B={I0(g)g=0,1,,n}.

Lemma 3.

For Problem P1 with pjs=pj+αs and ojs=oj+βs, an optimal schedule I exists in B.

Proof.

Suppose that IB and I=g. Then, for each scenario sS, (21)zsI=jICjs+jJIojs=z0I+gg+12αs+n-gβsz0I0g+gg+12αs+n-gβs=zsI0g.The proof is complete.

Theorem 4.

Both scenario cases of Problem P1 with pjs=pj+αs and ojs=oj+βs can be solved in O(kn2).

Proof.

By Lemma 3, a schedule IB with the minimum Z(I) is optimal. The values zs(Is) for all sS can be obtained in O(kn2) by the algorithm of Engels et al.  and Ghosh . The set B can be obtained in O(n2) by the algorithm of Choi and Chung . Since zs(I) can be calculated in O(n) for each IB and sS, the value Z(I) can be calculated in O(kn). Hence, the values Z(I) for all IB can be found in O(kn2).

Henceforth, we consider case (ii).

Lemma 5.

If pjs=p for each jJ and each sS and the ordered conditions are satisfied, then an optimal schedule I exists in (22)j,,nj=1,2,,n.

Proof.

Suppose that, in an optimal schedule I,

v<w;

vI and wI.

Let I- be a new schedule constructed by letting I-=I{w}{v}. Now, we will show that Z(I-)Z(I).

First, we consider the discrete scenario case. Then, it is observed that (23)zsI-zsIforeachsS.Thus, Lemma 5 holds.

Before considering the interval scenario case, we introduce the concept of the worst case scenario τ=(τ1,τ2,,τn) for I, defined as follows: let τ be the worst case scenario for I if (24)ZI=zτI-zτIτ.Furthermore, it is observed that the worst case scenario τ for I can be constructed as follows : (25)τj=ojL,forjIojU,forjI.If ovUowL, then, for each scenario sS, (26)zsI-zsI,which implies that Z(I-)Z(I). Thus, we assume that ovU>owL. Let τ and τ- be the worst case scenarios of I and I-, respectively. It is observed from (25) that (27)ovτ=ovL,owτ=owU,ovτ-=ovU,owτ-=owL,ojτ=ojτ-forjJv,w.Then, (28)zτI-zτ-I-=owU-ovU0.Consider the following three cases.

Jobs in {v,w} are processed in-house in Iτ-. Then, since zτ-(Iτ-)=zτ(Iτ-)zτ(Iτ), it is observed from (28) that Z(I-)Z(I).

Exactly one job in {v,w} is processed in-house in Iτ-. Since ovτ-=ovU>owτ-=owL, (29)vIτ-,wIτ-.Let I=Iτ-{w}{v}. Then, since zτ-(Iτ-)=zτ-(I)+owL-ovU=zτ(I)+owL-ovLzτ(I)zτ(Iτ), it is observed from (28) that Z(I-)Z(I).

No job in {v,w} is processed in-house in Iτ-. Since zτ-(Iτ-)=zτ(Iτ-)-(owU-ovU)+(owL-ovL)zτ(Iτ-)-(owU-ovU)zτ(Iτ)-(owU-ovU), it is observed from (28) that Z(I-)Z(I).

By cases  (i)–(iii), Z(I-)Z(I). Thus, by repeatedly applying the argument above, we can construct the optimal schedule satisfying Lemma 5 from I without an increase of the objective value. The proof is complete.

Theorem 6.

Both scenario cases of Problem P1 with pjs=p for each jJ and each sS and the ordered conditions can be solved in O(n2).

Proof.

This holds immediately from Lemma 5.

3. Problem P2

In this section, we show the discrete scenario case of Problem P2 remains NP-hard even if only one of the processing times and outsourcing costs is uncertain and we introduce a polynomially solvable case. Note that the complexity of the interval scenario case with certain processing times is open. Since the processing time of each job is identical for each scenario, in-house jobs are processed in nonincreasing order of wj in an optimal schedule. Thus, for notational simplicity, we use I and Cjs instead of (I,σ) and Cjs(σ), respectively, throughout this section.

3.1. NP-Hardness

In this subsection, we prove NP-hardness for the discrete scenario case of Problem P2.

Theorem 7.

The discrete scenario case of Problem P2 is NP-hard, even if the outsourcing cost is certain; that is, ojs=oj for each jJ.

Proof.

We reduce the even-odd partition problem known to be NP-complete , defined below, to Problem P2: given 2q positive integers {a1,a2,,a2q} with j=12qaj=2A, is there a subset Q such that jQaj=A and exactly one of the jobs in {2j-1,2j},j=1,2,,q, belongs to Q?

Given an instance of the even-odd partition problem, we can construct an instance of Problem P2 with two scenarios, as follows: set n=3q and J={1,2,,3q}. Let p1=0 and p2=2. For j=1,2,,q,

wj=M4q, and oj=M6q,

and, for j=1,2,,q,

wq+2j-1=(M2q-j+1+a2j-1)/(q+j),

wq+2j=(M2q-j+1+a2j)/(q+j),

oq+2j-1=M2q-j+1+a2j-1, and oq+2j=M2q-j+1+a2j,

where M>0 is a sufficiently large value. Clearly, this reduction can be done in polynomial time. Note the following:

The optimal schedule of scenario 1 is to process all jobs in-house. Thus, (30)z1I1=0.

The optimal schedule of scenario 2 is to process each job in {1,2,,q} in-house and to outsource all the other jobs. Thus, (31)z2I2=qq+1M4q+2j=1qMq+j+2A.

We now show that a schedule I exists such that (32)ZIj=1qMq+j+Aif and only if a solution to the even-odd partition problem exists.

Suppose that there exists a solution Q- to the even-odd partition problem. We can construct a schedule by letting I-={1,2,,q}{q+jjQ-}. Then, in a schedule I-, we have the following:

Exactly one of the jobs in {q+2j-1,q+2j},j=1,2,,q, is outsourced. Thus, under scenarios 1 and 2, the total outsourcing cost is (33)jJI-oj=2j-1Q-oq+2j-1+2jQ-oq+2j=j=1qM2q-j+1+jQ-aj=j=1qMq+j+A.

The total weighted completion time is, under scenario 1, (34)jI-wjp1=0and, under scenario 2, it is (35)jI-wjp2=2j=1qjM4q+22j-1Q-M2q-j+1+a2j-1+22jQ-M2q-j+1+a2j=qq+1M4q+2j=1qMq+j+2A.

Then, (36)ZI-=maxz1I--z1I1,z2I--z2I2=j=1qMq+j+A.

Suppose that a schedule I^ exists such that (37)ZI^j=1qMq+j+A.It is observed that jobs in {1,2,,q} should be processed in-house in I^ due to oj=M6q, j=1,2,,q; that is, {1,2,,q}I^. For notational simplicity, let Q^={jq+jI^}.

Claim. Exactly one of the jobs in {2j-1,2j},j=1,2,,q, is in Q^.

Proof. Let v be the smallest index such that both jobs in {2v-1,2v} are not in Q^, and let w be the smallest index such that both jobs in {2w-1,2w} are in Q^. If v or w does not exist, then let v= and w= for notational consistency. Consider the following two cases.

v<w.

In this case, exactly one of the jobs in {q+2j-1,q+2j},j=1,2,,v-1, is outsourced and both of the jobs in {q+2v-1,q+2v} are outsourced. Thus, (38)z1I^>j=1v-1M2q-j+1+2M2q-v+1.Since M2q-v+1>j=v+1qM2q-j+1+A, (39)ZI^z1I^-z1I1>j=1v-1M2q-j+1+2M2q-v+1>j=1qMq+j+A.This is a contradiction.

v>w.

In this case, exactly one of the jobs in {q+2j-1,q+2j},j=1,2,,w-1, is processed in-house, and both of the jobs in {q+2w-1,q+2w} are processed in-house. Thus, (40)z2I^>qq+1M4q+3j=1w-1M2q-j+1+4M2q-w+1.Thus, since M2q-w+1>3(j=w+1qM2q-j+1+A), (41)ZI^z2I^-z2I2>3j=1qM2q-j+1+3A-2j=1qMq+j+2A>j=1qMq+j+A.This is a contradiction. By cases (i) and (ii), the proof is complete.

By the above claim, we have (42)z1I^=2j-1Q^M2q-j+1+a2j-1+2jQ^M2q-j+1+a2j=j=1qMq+j+jQ^aj,z2I^=qq+1M4q+22j-1Q^M2q-j+1+a2j-1+2j-1Q^M2q-j+1+a2j-1+22jQ^M2q-j+1+a2j+2jQ^M2q-j+1+a2j.Since 2j-1Q^a2j-1+2j-1Q^a2j-1+22jQ^a2j+2jQ^a2j=2A, (43)z2I^=qq+1M4q+3j=1qM2q-j+1+2A+jQ^aj.Hence, (44)ZI^=j=1qMq+j+maxjQ^aj,jQ^aj.This implies that jQ^aj=A and Q^ is the solution to the even-odd partition problem.

Theorem 8.

The discrete scenario case of Problem P2 is NP-hard, even if the processing time of each job is identical under all scenarios; that is, pjs=1 for each jJ and each sS.

Proof.

Given an instance of the partition problem, we can construct an instance of Problem P2 with two scenarios, as follows: set n=2q and J={1,2,,2q}. For j=1,2,,q,

w2j-1=w2j=Mq-j+1,

p2j-11=p2j1=p2j-12=p2j2=1,

o2j-11=0 and o2j1=aj,

o2j-12=(4(j-1)+3)Mq-j+1 and o2j2=(4(j-1)+3)Mq-j+1-aj,

where M>0 is a sufficiently large value. Clearly, this reduction can be done in polynomial time. Since the proof is almost similar to that of Theorem 1, we omit it.

3.2. Polynomiality

In this subsection, we prove the polynomiality of the case with pjs=p and ojs=oj+βs for each jJ and each sS.

Using the following lemma, the interval scenario case is reduced to the discrete scenario case.

Lemma 9.

For the interval scenario case of Problem P2 with pjs=p and ojs=oj+βs, Z(I) is determined by a scenario in (45)βL,βU.

Proof.

Since the proof is almost similar to that of Lemma 2, we omit it.

Now, we focus on the discrete scenario case. Let I-0(g) be the schedule that minimizes z0(I-) among schedules with g in-house jobs, where z0(I-) is the total cost of I under βs=0. Let B-={I-0(g)g=0,1,,n}.

Lemma 10.

For Problem P2 with pjs=p and ojs=oj+βs, an optimal schedule I exists in B-.

Proof.

Since the proof is almost similar to that of Lemma 3, we omit it.

Theorem 11.

Both scenario cases of Problem P2 with pjs=p and ojs=oj+βs can be solved in O(n2).

Proof.

This holds immediately from Lemmas 9 and 10.

4. Concluding Remarks and Future Work

We considered the min–max regret version of a single-machine problem with an outsourcing option such that

the processing time or the outsourcing cost of each job is uncertain,

the cost for processing jobs in-house is expressed as the total weighted completion time.

The uncertainty was described as being of two types with a discrete scenario and an interval scenario. Since the deterministic version of the problem is known to be NP-hard, we considered two cases with identical weights or identical processing times for each scenario. For the first case, we proved the NP-hardness of its discrete scenario case, even if the processing time is certain. Furthermore, we presented two special structures for the processing time and outsourcing cost, each of which makes the first case polynomially solvable. For the second case, we proved the NP-hardness of its discrete scenario case, even if one of the outsourcing cost and the processing time is uncertain. Finally, we presented one special structure for the processing time and outsourcing cost that makes the second case polynomially solvable.

For Problems P1 and P2, it is observed that the complexities remain NP-hard, only if one of the processing time and the outsourcing cost is uncertain. However, their complexities move to the polynomially solvable class, if the impact of uncertainty on each job under each scenario is the same, which is described by pjs=pj+αs and ojs=oj+βs. Thus, our description for the impact of uncertainty may be the starting point to find additional polynomially solvable cases.

For future work, it would be interesting to consider the following:

Developing the algorithm for Problem P whose performance is verified through analyzing the approximability or conducting the numerical experiments.

Analysis of computational complexity for the interval scenario cases of Problems P1 and P2 when the processing time is certain.

Analysis of the min–max version of Problems P1 and P2 to find an optimal schedule (I,σ) for minimizing (46)maxsSzsI,σ.

Note that the discrete scenario case of the min–max version of Problem P1 is NP-hard, even if only the outsourcing cost is uncertain. The reduction from the partition problem is as follows: For j=1,2,,q,

p2j-1=p2j=Mj,

o2j-11=(q-j+3/2)Mj+aj and o2j1=(q-j+3/2)Mj,

o2j-12=(q-j+3/2)Mj and o2j2=(q-j+3/2)Mj+aj,

K=j=1q(2q-2j+5/2)Mj+A/2,

where M>0 is a sufficiently large value. Since the proof is very similar to that of Theorem 1, we omit it.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This study was financially supported by the research fund of Chungnam National University in 2016.

Daniels R. L. Kouvelis P. Robust scheduling to hedge against processing time uncertainty in single-stage production Management Science 1995 41 2 363 376 10.1287/mnsc.41.2.363 ZBL0832.90050 2-s2.0-0029254356 Kouvelis P. Yu G. Robust Discrete Optimization and Its Applications 1997 14 Dordrecht, Netherlands Kluwer Academic Nonconvex Optimization and Its Applications 10.1007/978-1-4757-2620-6 MR1480918 Aissi H. Bazgan C. Vanderpooten D. Min-max and min-max regret versions of combinatorial optimization problems: a survey European Journal of Operational Research 2009 197 2 427 438 10.1016/j.ejor.2008.09.012 MR2509273 ZBL1159.90472 2-s2.0-60749104714 Vickson R. G. Choosing the job sequence and processing times to minimize total processing plus flow cost on a single machine Operations Research 1980 28 5 1155 1167 10.1287/opre.28.5.1155 2-s2.0-0019058538 Shabtay D. Gaspar N. Kaspi M. A survey on offline scheduling with rejection Journal of Scheduling 2013 16 1 3 28 10.1007/s10951-012-0303-z MR3030200 ZBL1297.90058 2-s2.0-84874114189 Choi B.-C. Chung K. Min-max regret version of a scheduling problem with outsourcing decisions under processing time uncertainty European Journal of Operational Research 2016 252 2 367 375 10.1016/j.ejor.2015.12.046 MR3466705 ZBL1346.90333 2-s2.0-84959909788 Engels D. W. Karger D. R. Kolliopoulos S. G. Sengupta S. Uma R. N. Wein J. Techniques for scheduling with rejection Journal of Algorithms. Cognition, Informatics and Logic 2003 49 1 175 191 10.1016/S0196-6774(03)00078-6 MR2027063 ZBL1067.68024 2-s2.0-0242667531 Ghosh J. B. Job selection in a heavily loaded shop Computers and Operations Research 1997 24 2 141 145 10.1016/S0305-0548(96)00045-7 ZBL0893.90089 2-s2.0-0031075465 Conde E. A 2-approximation for minmax regret problems via a mid-point scenario optimal solution Operations Research Letters 2010 38 4 326 327 10.1016/j.orl.2010.03.002 MR2647247 ZBL1197.90337 2-s2.0-77955292984 Conde E. On a constant factor approximation for minmax regret problems using a symmetry point scenario European Journal of Operational Research 2012 219 2 452 457 10.1016/j.ejor.2012.01.005 MR2888120 ZBL1244.90241 2-s2.0-84857043584 Kasperski A. Zieliński P. A 2-approximation algorithm for interval data minmax regret sequencing problems with the total flow time criterion Operations Research Letters 2008 36 3 343 344 10.1016/j.orl.2007.11.004 MR2424459 2-s2.0-43049090878 Lebedev V. Averbakh I. Complexity of minimizing the total flow time with interval data and minmax regret criterion Discrete Applied Mathematics 2006 154 15 2167 2177 10.1016/j.dam.2005.04.015 MR2257820 2-s2.0-33747042784 Garey M. R. Johnson D. S. Computers and Intractability: A Guide to the Theory of NP-Completeness 1979 New York, NY, USA W. H. Freeman MR519066 Kasperski A. Zieliński P. An approximation algorithm for interval data minmax regret combinatorial optimization problems Information Processing Letters 2006 97 5 177 180 10.1016/j.ipl.2005.11.001 MR2195214 2-s2.0-30344473284