Juta Pichitlamken

Markov chain models of a telephone call center with call blending

Motivated by a Bell Canada call center operating in blend mode, we consider a system with two types of traffic and two types of agents. Outbound calls are served only by blend agents, whereas inbound calls can be served by either inbound-only or blend agents. Inbound callers may balk or abandon. There are several performance measures of interest, including the rate of outbound calls and the proportion of inbound calls waiting more than some fixed number of seconds. We present a collection of continuous-time Markov chain (CTMC) models which capture many real-world characteristics while maintaining parsimony that results in fast computation. We discuss and explore the tradeoffs between model fidelity and efficacy and compare our different CTMC models with a realistic simulation model of a Bell Canada call center, used as a benchmark.

A sequential procedure for neighborhood selection-of-the-best in optimization via simulation

We propose a fully sequential indifference-zone selection procedure that is specifically for use within an optimization-via-simulation algorithm when simulation is costly, and partial or complete information on solutions previously visited is maintained. Sequential Selection with Memory guarantees to select the best or near-best alternative with a user-specified probability when some solutions have already been sampled, their previous samples are retained, and simulation outputs are i.i.d. normal. For the case when only summary information on solutions is retained, we derive a modified procedure. We illustrate how our procedures can be applied to optimization-via-simulation problems and compare its performance with other methods by numerical examples.

A combined procedure for optimization via simulation

We propose an optimization-via-simulation algorithm for use when the performance measure is estimated via a stochastic, discrete-event simulation, and the decision variables may be subject to deterministic linear integer constraints. Our approach-which consists of a global guidance system, a selection-of-the-best procedure, and local improvement-is globally convergent under very mild conditions.

Selection-of-the-best procedures for optimization via simulation

We propose fully sequential indifference-zone selection procedures that are specifically for use within an optimization-via-simulation algorithm when simulation is costly and partial or complete information on solutions previously visited is maintained. Sequential Selection with Memory guarantees to select the best or near-best alternative with a user-specified probability when some solutions have already been sampled and their previous samples are retained. For the case when only summary information is retained, we derive a modified procedure. We illustrate how our procedure can be applied to optimization-via-simulation problems and compare its performance with other methods by numerical examples.

A Generic Discrete-Event Simulation Model for Outpatient Clinics in a Large Public Hospital

The orthopedic outpatient department (OPD) ward in a large Thai public hospital is modeled using Discrete-Event Stochastic (DES) simulation. Key Performance Indicators (KPIs) are used to measure effects across various clinical operations during different shifts throughout the day. By considering various KPIs such as wait times to see doctors, percentage of patients who can see a doctor within a target time frame, and the time that the last patient completes their doctor consultation, bottlenecks are identified and resource-critical clinics can be prioritized. The simulation model quantifies the chronic, high patient congestion that is prevalent amongst Thai public hospitals with very high patient-to-doctor ratios. Our model can be applied across five different OPD wards by modifying the model parameters. Throughout this work, we show how DES models can be used as decision-support tools for hospital management.

High performance spreadsheet simulation on a desktop grid

We present a proof-of-concept prototype for high performance spreadsheet simulation called S3. Our goal is to provide a user-friendly, yet computationally powerful simulation environment for end users. Our approach is to add power of parallel computing on Windows-based desktop grid into popular Excel models. We show that, by using standard Web services and service-oriented architecture (SOA), one can build a fast and efficient system on a desktop grid for simulation. The complexity of parallelism can be hidden from users through a well-defined computation template. This work also demonstrates that a massive computing power can be harvested by linking off-the-shelf office PCs into a desktop grid for simulation. The experimental results show that the prototype system is highly scalable. In the best case, the execution time can be reduced 13.6 times using 16 desktop PCs; the simulation time is dramatically reduced from 200 minutes to 14 minutes.

Applying Web service and Windows clustering for high volume risk analysis

We present the development of a distributed system to calculate the Value at Risk (VaR) measure when a large number of users are presented. A scalable architecture based on Windows clustering and Web services is proposed. In addition, we develop a load balancing algorithm to distribute the workload among the compute nodes in the Windows cluster. The experimental results show that our system can substantially speed up the VaR calculation. In addition, it offers a good scalability. This work provides an example of how to deploy a standard Web service and Windows clustering technology to offer a cost-effective and scalable solution for demanding financial applications in practice

Properties of Optimal-Weighted Flowtime Policies with a Makespan Constraint and Set-up Times

We characterize optimal policies for the problem of allocating a single server to a set of jobs fromN families. Each job is an instance of demand for an item and is associated with a family, a holding cost rate, and a mean processing time. Set-up times are required to switch from one family to another, but are not required to switch within a family. We consider the case in which the order of jobs within the family is unconstrained, and a variation in which the order is fixed. The optimization is with respect to the weighted flowtime, and we treat problems both with and without a makespan-constraint. Practical examples based on this model are described. We partially characterize an optimal policy by means of a Gittins rewardrate index and a similar switching index derived from multi-armed bandit theory. For deterministic problems with a makespan constraint, we present an optimization algorithm for the special case of two families and at most three set-ups . Without a makespan constraint and without preemption, we prove that our analysis of a deterministic model extends to stochastic set-up and processing times without loss of optimality. Managerial insights based on our technical results are provided.

Nelder-Mead Method with Local Selection Using Neighborhood and Memory for Stochastic Optimization

We consider the Nelder-Mead (NM) simplex algorithm for optimization of discrete-event stochastic simulation models. We propose new modifications of NM to reduce computational time and to improve quality of the estimated optimal solutions. Our means include utilizing past information of already seen solutions, expanding search space to their neighborhood and using adaptive sample sizes. We compare performance of these extensions on six test functions with 3 levels of random variations. We find that using past information leads to reduction of computational efforts by up to 20%. The adaptive modifications need more resources than the non-adaptive counterparts for up to 70% but give better-quality solutions. We recommend the adaptive algorithms with using memory with or without neighborhood structure.

Assessing risk in a job schedule: integrating a scheduling heuristic and a simulation model to a spreadsheet

Deterministic scheduling algorithms are often applied to problems in stochastic settings perhaps because they are already hard to solve even without considering stochastic characteristics. We are interested in assessing the measure of risk in performance measures (e.g., makespan) when these algorithms are used in probabilistic environment. We design an easy-to-use Microsoft Excel program that integrates a Visual Basic Application (VBA) subroutine which performs scheduling procedures, with an Arena simulation model that imitates the stochastic production environment. Our program suggests a job schedule, its associated performance measures and the corresponding prediction intervals. At the moment, we only consider the m-machine permutation flowshop problem with the makespan (or completion time) objective.