Benchmarks by Classical Instances

Speed Comparison

Optimality Proven Instances

Instance Name	Cplex	Gurobi	AutoRoster	ScheduleNurse3
QMC-1	2.95/2.25=1.3	2.25/2.25=1	-	8.5/2.25=3.8
SINTEF	1.89/0.78=2.4	0.78/0.78=1	9/0.78=11.5	1.15/0.78=1.5
ikegami-3Shift-DATA1.2	-	695/5.66=122.8	-	5.66/5.66=1
ikegami-3Shift-DATA1.1	6606/7.155=923.3	416/7.155=58.1	-	7.155/7.155=1
ikegami-3Shift-DATA1	1838/4=459.5	285/4=71.3	-	4/4=1
ikegami-2Shift-DATA1	9.23/0.14=65.9	0.14/0.14=1	11/0.14=78.6	1.94/0.14=13.9
GPOST-B	227/34=6.7	161/34=4.7	40/34=1.2	34/34=1
GPOST	124/2.8=44.3	22/2.8=7.9	17/2.8=6.1	2.8/2.8=1
Valouxis-1	-	-	-	37/37=1
WHPP	-	4853/4=1213.3	17/4=4.3	4/4=1
BCDT-Sep	-	-	-	140/140=1

Optimal Objective Reached Instances

Instance Name	Cplex	Gurobi	AutoRoster	ScheduleNurse3
QMC-1	2.95/2.25=1.3	2.25/2.25=1	140/2.25=62.2	8.5/2.25=3.8
SINTEF	1.89/0.78=2.4	0.78/0.78=1	9/0.78=11.5	1.146/0.78=1.5
ikegami-3Shift-DATA1.2	2573/4=643.3	184/4=46	-	4/4=1
ikegami-3Shift-DATA1.1	6606/3.94=1676.6	175/3.94=44.4	-	3.94/3.94=1
ikegami-3Shift-DATA1	1200/4=300	285/4=71.3	300/4=75	4/4=1
ikegami-2Shift-DATA1	9.23/0.14=65.9	0.14/0.14=1	11/0.14=78.6	1.94/0.14=13.9
GPOST-B	130/2.5=52	61/2.5=24.4	40/2.5=16	2.5/2.5=1
GPOST	124/2.3=53.9	22/2.3=9.6	17/2.3=7.4	2.3/2.3=1
Valouxis-1	663/3.91=170	224/3.91=57.3	9/3.91=2.3	3.91/3.91=1
WHPP	-	4853/4=1213.3	17/4=4.3	4/4=1
BCDT-Sep	-	-	-	140/140=1

Time - Number of Instances proven optimality

No.	Instance Name
1	Millar-2Shift-DATA1.1
2	Millar-2Shift-DATA1
3	Ozkarahan
4	Musa
5	Azaiez
6	QMC-1
7	LLR
8	SINTEF
9	ikegami-3Shift-DATA1.2
10	ikegami-3Shift-DATA1.1
11	ikegami-3Shift-DATA1
12	ikegami-2Shift-DATA1
13	GPOST-B
14	BCV-4.13.1
15	GPOST
16	Valouxis-1
17	WHPP
18	BCDT-Sep

Time - Number of Instances reached optimal objective

No.	Instance Name
1	Millar-2Shift-DATA1.1
2	Millar-2Shift-DATA1
3	Ozkarahan
4	Musa
5	Azaiez
6	QMC-1
7	LLR
8	SINTEF
9	ikegami-3Shift-DATA1.2
10	ikegami-3Shift-DATA1.1
11	ikegami-3Shift-DATA1
12	ikegami-2Shift-DATA1
13	GPOST-B
14	BCV-4.13.1
15	GPOST
16	Valouxis-1
17	WHPP
18	BCDT-Sep

Environment

Solver	Version	Machine
Gurobi	Gurobi Optimizer version 9.5.0 build v9.5.0rc5	NEOS SERVER
AutoRoster	RosterViewerDemo4.3.5 Branch and Price	Ryzen 5800X 64GB
Cplex	IBM(R) ILOG(R) CPLEX(R) Interactive Optimizer 20.1.0.0	NEOS SERVER
Schedule Nurse3	Algorithm3	Ryzen 5800X 64GB

References

1. Asta, S., Özcan, E., and Curtois, T. A tensor based hyper-heuristic for nurse rostering. Knowledge-based systems, 2016. 98: p. 185-199.

2. Burke E.K. and T. Curtois. New Approaches to Nurse Rostering Benchmark Instances. European Journal of Operational Research, 2014. 237(1): p. 71-81. pdf.

3. Solos, Ioannis P., Ioannis X. Tassopoulos and Grigorios N. Beligiannis. A Generic Two-Phase Stochastic Variable Neighborhood Approach for Effectively Solving the Nurse Rostering Problem. Algorithms, 2013. 6: p. 278-308.