Sameh Al-Shihabi

A hybrid of Nested Partition, Binary Ant System, and Linear Programming for the multidimensional knapsack problem

This work presents a hybrid algorithm of Nested Partition (NP), Binary Ant System (BAS), and Linear Programming (LP) to solve the multidimensional knapsack problem (MKP). The hybrid NP+BAS+LP algorithm takes advantage of the global search strategies of the NP algorithm; the ability of BAS to quickly generate good solutions and incorporates information obtained from solving a LP relaxation of the MKP to help guide the search. It thus incorporates both the lower bounds (LB), found by the BAS, and the upper bounds (UB), found by solving the relaxed LP, into the search by embedding both in the NP framework. An adjustable computation budget is implemented where the number of samples increases if the LB and the UB point to different promising subregions. The proposed hybrid is compared to state-of-the-art solution techniques and is found to be one of the best algorithms in terms of the quality of solutions obtained and CPU time requirements.

Nested Partitioning for the Minimum Energy Broadcast Problem

The problem of finding the broadcast scheme with minimum power consumption in a wireless ad-hoc network is NP-hard. This work presents a new hybrid algorithm to solve this problem by combining Nested Partitioning with Local Search and Linear Programming. The algorithm is benchmarked by solving instances with 20 and 50 nodes where results are compared to either optimum or best results found by an IP solver. In these instances, the proposed algorithm was able to find optimal and near optimal solutions.

Reducing waiting time at an emergency department using design for Six Sigma and discrete event simulation

Design for Six Sigma (DFSS) has been implemented in different industries as a methodology to design or redesign processes. In this paper, DFSS is used to develop a triage process for an emergency department (ED) at a Jordanian hospital. Different performance measures, such as length of stay (LOS) and waiting time (WT), are employed to evaluate the hospitals ED performance before and after the triage process. Discrete event simulation (DES) models were developed using ProModel software. The models have been verified and validated. The results indicate that LOS will be reduced by 34% and WT by 61% after the triage system is implemented, without any additional staff. Moreover, as a result of the triage process, the WT sigma level is improved from 0.66 to 5.18, and the LOS sigma level is improved from 0.58 to 3.09.

A statistical study employing agent-based modeling to estimate the effects of different warehouse parameters on the distance traveled in warehouses

Abstract In manually operated warehouses, the travel distance of the order picker has profound effects on the warehouse cost and efficiency. Estimating this distance is difficult because the warehouse environment is a stochastic one, affected by a great number of parameters. Therefore, we present a comprehensive statistical study to assess how the different warehouse parameters and their interactions affect the travel distance. To estimate the travel distance, we simulate the different designs using agent-based modeling (ABM). Having 324 different designs, ABM has enabled us to build one computer model to simulate all the cases. The study shows that having one cross aisle only and using a class-based storage policy decreases the travel distance. Moreover, the results obtained show that choosing the best routing policy depends on the warehouse layout, which proves the importance of considering the interactions among the different parameters.

An improved hybrid algorithm for the set covering problem

Discussing a number of weak points of a previous algorithm to solve the set covering problem (SCP).Developing a new hybrid algorithm that has the best performance among all meta-heuristics to solve the SCP.Proposing a new mechanism to update the pheromone trails limits in a Max-Min Ant System (MMAS).Using a simple normalizing step to deal with possible ranges of heuristic information.Very low computation times. The state-of-the-art ant colony optimization (ACO) algorithm to solve large scale set covering problems (SCP) starts by solving the Lagrangian dual (LD) problem of the SCP to obtain quasi-optimal dual values. These values are then exploited by the ACO algorithm in the form of heuristic estimates. This article starts by discussing the complexity of this approach where a number of new parameters are introduced to escape local optimums and normalize the heuristic values. To avoid these complexities, we propose a new hybrid algorithm that starts by solving the linear programming (LP) relaxation of the SCP. This solution is used to eliminate unnecessary columns, and to estimate the heuristic information. To generate solutions, we use a Max-Min Ant System (MMAS) algorithm that employs a novel mechanism to update the pheromone trail limits to maintain a predetermined exploration rate. Computational experiments on different sets of benchmark instances prove that our proposed algorithm can be considered the new state-of-the-art meta-heuristic to solve the SCP.

Backtracking Ant System for the Traveling Salesman Problem

In this work, we adopt the concept of backtracking from the Nested Partition (NP) algorithm and apply it to the Max-Min Ant System (MMAS) to solve the Traveling Salesman Problem (TSP). A new type of ants that is called backtracking ants (BA) is used to challenge a subset of the solution feasible space that is expected to have the global optimum solution. The size of this subset is decreased if the BAs find a better solution out of this subset or increased if the BAs fail in their challenge. The BAs don’t have to generate full tours like previous ant systems, which leads to a considerable reduction in the computation effort. A computational experiment is conducted to check the validity of the proposed algorithm.

A max-min ant system for the finance-based scheduling problem

Abstract Construction contractors depend on bank overdrafts to finance their expenses; however, these overdrafts cannot exceed an imposed Credit Line (CL). The Finance-Based Scheduling Problem (FBSP) is about scheduling activities without exceeding the CL. In this paper, we provide a more eloquent formulation of the FBSP and list its different variants. Three Max-Min Ant System (MMAS) algorithms, which use different heuristic information when generating solutions, are then developed to solve the FBSP. To test the MMAS algorithms, we generate 60 instances that are used to tune the MMAS algorithms and then use these algorithms to solve the generated instances. The found solutions are compared with the best bounds found using a Branch and Bound (BB moreover, the comparison shows that using the number of successors as heuristic outperformed other heuristics. Furthermore, the MMAS algorithm outperformed other meta-heuristics that use repair operators or penalize infeasible solutions in terms of computation time while having comparable solution values.

Utilizing six sigma to improve the processing time: a simulation study at an emergency department

Abstract Waiting time (WT) at the emergency department (ED) is a global concern, emerging evidence indicates that a wait for care delivery may result in adverse patient outcomes. Discrete-event simulation model has been developed to redesign the existing ED based upon several inputs (i.e. historical data, staff survey, and interviews). Furthermore, a newly developed simulation model was proposed, verified, and validated using triaged management system based upon Manchester triage system. The simulation study was performed as a part of design for six sigma project to create the proposed triage process. The proposed model resulted in reducing WT by 61% and length of stay (LOS) by 34%. In return, the sigma level was improved from 0.66 to 5.18 and from 0.58 to 3.09 for WT and LOS, respectively.

Selecting the best supplier based on a multi-criteria taguchi loss function: a simulation optimization approach

Minimum price is not the only objective that companies pursue when sourcing their materials. Selecting the best supplier entails looking for the best quality as well as the most reliable delivery. This work suggests a Multi-Criteria objective function that linearly aggregates a number of Taguchi loss functions, which represent the criteria of price, quality, and delivery. We initially recommend a framework to represent the market and then generate test data to represent the different market scenarios. We introduce randomness into this framework in order to achieve a highly realistic assumption. This study then employs the Optimal Computation Budget Allocation (OCBA) algorithm to choose the best supplier. OCBA solutions are benchmarked against the deterministic solution to check OCBAs ability to find the optimal solution. OCBA solutions are also compared to an Equal Allocation (EA) algorithm to verify their effectiveness in terms of minimizing the costs of sampling.

A reinforcement learning algorithm to minimize the mean tardiness of a single machine with controlled capacity

In this work, we consider the problem of scheduling arriving jobs to a single machine where the objective is to minimize the mean tardiness. The scheduler has the option of reducing the processing time by half through the employment of an extra worker for an extra cost per job (setup cost). The scheduler can also choose from a number of dispatching rules. To find a good policy to be followed by the scheduler, we implemented a lambda-SMART algorithm to do an on-line optimization for the studied system. The found policy is only optimal with respect to the state representation and set of actions available, however, we believe that the developed policies are easy to implement and would result in considerable savings as shown by the numerical experiments conducted