Aldy Gunawan

Orienteering Problem: A Survey of Recent Variants, Solution Approaches and Applications

The Orienteering Problem (OP) has received a lot of attention in the past few decades. The OP is a routing problem in which the goal is to determine a subset of nodes to visit, and in which order, so that the total collected score is maximized and a given time budget is not exceeded. A number of typical variants has been studied, such as the Team OP, the (Team) OP with Time Windows and the Time Dependent OP. Recently, a number of new variants of the OP was introduced, such as the Stochastic OP, the Generalized OP, the Arc OP, the Multi-agent OP, the Clustered OP and others. This paper focuses on a comprehensive and thorough survey of recent variants of the OP, including the proposed solution approaches. Moreover, the OP has been used as a model in many different practical applications. The most recent applications of the OP, such as the Tourist Trip Design Problem and the mobile-crowdsourcing problem are discussed. Finally, we also present some promising topics for future research.

Fine-Tuning algorithm parameters using the design of experiments approach

Optimizing parameter settings is an important task in algorithm design. Several automated parameter tuning procedures/configurators have been proposed in the literature, most of which work effectively when given a good initial range for the parameter values. In the Design of Experiments (DOE), a good initial range is known to lead to an optimum parameter setting. In this paper, we present a framework based on DOE to find a good initial range of parameter values for automated tuning. We use a factorial experiment design to first screen and rank all the parameters thereby allowing us to then focus on the parameter search space of the important parameters. A model based on the Response Surface methodology is then proposed to define the promising initial range for the important parameter values. We show how our approach can be embedded with existing automated parameter tuning configurators, namely ParamILS and RCS (Randomized Convex Search), to tune target algorithms and demonstrate that our proposed methodology leads to improvements in terms of the quality of the solutions.

Master Physician Scheduling Problem

We study a real-world problem arising from the operations of a hospital service provider, which we term the master physician scheduling problem. It is a planning problem of assigning physicians’ full range of day-to-day duties (including surgery, clinics, scopes, calls, administration) to the defined time slots/shifts over a time horizon, incorporating a large number of constraints and complex physician preferences. The goals are to satisfy as many physicians’ preferences and duty requirements as possible while ensuring optimum usage of available resources. We propose mathematical programming models that represent different variants of this problem. The models were tested on a real case from the Surgery Department of a local government hospital, as well as on randomly generated problem instances. The computational results are reported together with analysis on the optimal solutions obtained. For large-scale instances that could not be solved by the exact method, we propose a heuristic algorithm to generate good solutions.

An Iterated Local Search Algorithm for Solving the Orienteering Problem with Time Windows

The Orienteering Problem with Time Windows (OPTW) is a variant of the Orienteering Problem (OP). Given a set of nodes including their scores, service times and time windows, the goal is to maximize the total of scores collected by a particular route considering a predefined time window during which the service has to start. We propose an Iterated Local Search (ILS) algorithm to solve the OPTW, which is based on several LocalSearch operations, such as swap, 2-opt, insert and replace. We also implement the combination between AcceptanceCriterion and Perturbation mechanisms to control the balance between diversification and intensification of the search. In Perturbation, Shake strategy is introduced. The computational results obtained by our proposed algorithm are compared against optimal solutions or best known solution values obtained by state-of-the-art algorithms. We show experimentally that our proposed algorithm is effective on well-known benchmark instances available in the literature. It is also able to improve the best known solution of some benchmark instances.

Patrol scheduling in urban rail network

This paper presents the problem of scheduling security teams to patrol a mass rapid transit rail network of a large urban city. The main objective of patrol scheduling is to deploy security teams to stations of the network at varying time periods subject to rostering as well as security-related constraints. We present several mathematical programming models for different variants of this problem. To generate randomized schedules on a regular basis, we propose injecting randomness by varying the start time and break time for each team as well as varying the visit frequency and visit time for each station according to their reported vulnerability. Finally, we present results for the case of Singapore mass rapid transit rail network and synthetic instances.

Well-tuned algorithms for the Team Orienteering Problem with Time Windows

The Team Orienteering Problem with Time Windows (TOPTW) is the extension of the Orienteering Problem (OP) where each node is limited by a predefined time window during which the service has to start. The objective of the TOPTW is to maximize the total collected score by visiting a set of nodes with a limited number of paths. We propose two algorithms, Iterated Local Search and a hybridization of Simulated Annealing and Iterated Local Search (SAILS), to solve the TOPTW. As indicated in multiple research works on algorithms for the OP and its variants, determining appropriate parameter values in a statistical way remains a challenge. We apply Design of Experiments, namely factorial experimental design, to screen and rank all the parameters thereby allowing us to focus on the parameter search space of the important parameters. The proposed algorithms are tested on benchmark TOPTW instances. We demonstrate that well-tuned ILS and SAILS lead to improvements in terms of the quality of the solutions. More precisely, we are able to improve 50 best known solution values on the available benchmark instances.

Hybrid metaheuristics for solving the quadratic assignment problem and the generalized quadratic assignment problem

This paper presents a hybrid metaheuristic for solving the Quadratic Assignment Problem (QAP). The proposed algorithm involves using the Greedy Randomized Adaptive Search Procedure (GRASP) to construct an initial solution, and then using a hybrid Simulated Annealing and Tabu Search (SA-TS) algorithm to further improve the solution. Experimental results show that the hybrid metaheuristic is able to obtain good quality solutions for QAPLIB test problems within reasonable computation time. The proposed algorithm is extended to solve the Generalized Quadratic Assignment Problem (GQAP), with an emphasis on modelling and solving a practical problem, namely an examination timetabling problem. We found that the proposed algorithm is able to perform better than the standard SA algorithm does.

Real-world parameter tuning using factorial design with parameter decomposition

In this paper, we explore the idea of improving the efficiency of factorial design for parameter tuning of metaheuristics. In a standard full factorial design, the number of runs increases exponentially as the number of parameters. To reduce the parameter search space, one option is to first partition parameters into disjoint categories. While this may be done manually based on user guidance, an automated approach proposed in this paper is to apply a fractional factorial design to partition parameters based on their main effects where each partition is then tuned independently. With a careful choice of fractional design, our approach yields a linear rather than exponential run time performance with respect to the number of parameters. We empirically evaluate our approach for tuning a simulated annealing algorithm that solves an industry spares inventory optimization problem. We show that our proposed methodology leads to improvements in terms of the quality of solutions when compared to a pure application of an automated parameter tuning configurator ParamILS.

Designing and Comparing Multiple Portfolios of Parameter Configurations for Online Algorithm Selection

Algorithm portfolios seek to determine an effective set of algorithms that can be used within an algorithm selection framework to solve problems. A limited number of these portfolio studies focus on generating different versions of a target algorithm using different parameter configurations. In this paper, we employ a Design of Experiments (DOE) approach to determine a promising range of values for each parameter of an algorithm. These ranges are further processed to determine a portfolio of parameter configurations, which would be used within two online Algorithm Selection approaches for solving different instances of a given combinatorial optimization problem effectively. We apply our approach on a Simulated Annealing-Tabu Search (SA-TS) hybrid algorithm for solving the Quadratic Assignment Problem (QAP) as well as an Iterated Local Search (ILS) on the Travelling Salesman Problem (TSP). We also generate a portfolio of parameter configurations using best-of-breed parameter tuning approaches directly for the comparison purpose. Experimental results show that our approach lead to improvements over best-of-breed parameter tuning approaches.

ADOPT: Combining parameter tuning and Adaptive Operator Ordering for solving a class of Orienteering Problems

Abstract Two fundamental challenges in local search based metaheuristics are how to determine parameter configurations and design the underlying Local Search (LS) procedure. In this paper, we propose a framework in order to handle both challenges, called ADaptive OPeraTor Ordering (ADOPT). In this paper, The ADOPT framework is applied to two metaheuristics, namely Iterated Local Search (ILS) and a hybridization of Simulated Annealing and ILS (SAILS) for solving two variants of the Orienteering Problem: the Team Dependent Orienteering Problem (TDOP) and the Team Orienteering Problem with Time Windows (TOPTW). This framework consists of two main processes. The Design of Experiment (DOE) process, which is based on a 2 k factorial design, determines important parameters to tune and the best configuration for those parameters. The ADOPT process accommodates a reinforcement learning mechanism (based on Learning Automata) that calculates the probability of selecting an operator of LS. The probability values would be used to generate a sequence/order of operators for the next LS iteration, based on three different ordering strategies: rank-based, random and fitness proportionate selections. Our computational results show the superiority of the ADOPT framework with the fitness proportionate selection strategy against other ordering strategies in solving benchmark instances. In general, SAILS with the fitness proportionate selection strategy is competitive and comparable to the state-of-the-art algorithms. The proposed framework is able to improve the performances of both ILS and SAILS by discovering 11 new best known solutions of the benchmark TOPTW instances.