Angela Hsiang-Ling Chen

Application of a Hybrid Ant Colony Optimization for the Multilevel Thresholding in Image Processing

Our study proposes a hybrid optimization scheme based on an ant colony optimization algorithm with the Otsu method to render the optimal thresholding technique more applicable and effective. The properties of discriminate analysis in Otsu’s method are to analyze the separability among the gray levels in the image. The ACO-Otsu algorithm, a non-parametric and unsupervised method, is the first-known application of ACO to automatic threshold selection for image segmentation. The experimental results show that the ACO-Otsu efficiently speed up the Otsu’s method to a great extent at multi-level thresholding, and that such method can provide better effectiveness at population size of 20 for all given image types at multi-level thresholding in this study.

An artificial bee colony algorithm for the cardinality-constrained portfolio optimization problems

When conventional methods have become insufficient to handle computationally complicated problems, nature-inspired optimization methods were proposed and widely accepted in recent years. In this work, we investigate the trade-off between risk and return in a cardinality-constrained portfolio optimization problem and applied an artificial bee colony (ABC) method as the solution approach. It would be the first attempt of ABC on this application. The proposed ABC algorithm employs a hybrid encoding that mixes integer and real variables to fulfill the characteristic of the portfolio optimization problem. The generation of solutions involves three groups of bees: employed bees, onlooker and scouts that balance the effects of exploration and exploitation. The study tests the performance of the proposed ABC algorithm on four global stock market indexes provided by the OR-Library. Computational results of ABC are compared with simulated annealing (SA), tabu search (TS), and variable neighborhood search (VNS) methods in the literature. Evidences indicate that ABC performs better in terms of diversity, convergence, and effectiveness among all three test data sets; therefore, ABC demonstrates its potential on portfolio optimization.

A memetic algorithm for maximizing net present value in resource-constrained project scheduling problem

In this study, we develop a model that considers monetary issues in resource-constrained environments, and involves scheduling project activities to maximize net present value. This problem is recognized as the ldquoresource-constrained project scheduling problem with discounted cash flows (RCPSPDCF),rdquo. which is strongly NP-hard. All resources considered are both types of renewable and nonrenewable; the duration of each activity depends on the amount of resources allocated to its execution. Efforts are made by considering a two-stage method applying mode selection rules at the first stage and the memetic algorithm at the second stage. Results are shown in a comparative study which demonstrates the effectiveness of using memetic algorithm in maximizing project net present value; as well as, a combination of mode selection rules which provide a high probability of giving the best solution.

Artificial Bee Colony for workflow scheduling

Cloud computing is the provision of computing resource services from which users can obtain resources via network to tackle their demands. In recent years, with fast growing information technology, more users apply this service; as a result, the demand has increased dramatically. In addition, most of the complex tasks are represented by workflow and executed in the cloud. Therefore, as service providers face this increasing demand, how to schedule the workflow and reduce the response time becomes a critical issue. This research integrates the concept of project scheduling with the workflow scheduling problem to formulate a mathematical model, which expects to minimize the total completion time. Two Artificial Bee Colony algorithms are proposed to solve the workflow scheduling optimization problem. The performance of ABC is compared with the optimal solutions obtained by Gurobi optimizer on the instance containing different sizes of workflows. The results show that ABC can be considered a practical method for complicated workflow scheduling problems in the cloud computing environment.

An Entropy-Based Upper Bound Methodology for Robust Predictive Multi-Mode RCPSP Schedules

Projects are an important part of our activities and regardless of their magnitude, scheduling is at the very core of every project. In an ideal world makespan minimization, which is the most commonly sought objective, would give us an advantage. However, every time we execute a project we have to deal with uncertainty; part of it coming from known sources and part remaining unknown until it affects us. For this reason, it is much more practical to focus on making our schedules robust, capable of handling uncertainty, and even to determine a range in which the project could be completed. In this paper we focus on an approach to determine such a range for the Multi-mode Resource Constrained Project Scheduling Problem (MRCPSP), a widely researched, NP-complete problem, but without adding any subjective considerations to its estimation. We do this by using a concept well known in the domain of thermodynamics, entropy and a three-stage approach. First we use Artificial Bee Colony (ABC)—an effective and powerful meta-heuristic—to determine a schedule with minimized makespan which serves as a lower bound. The second stage defines buffer times and creates an upper bound makespan using an entropy function, with the advantage over other methods that it only considers elements which are inherent to the schedule itself and does not introduce any subjectivity to the buffer time generation. In the last stage, we use the ABC algorithm with an objective function that seeks to maximize robustness while staying within the makespan boundaries defined previously and in some cases even below the lower boundary. We evaluate our approach with two different benchmarks sets: when using the PSPLIB for the MRCPSP benchmark set, the computational results indicate that it is possible to generate robust schedules which generally result in an increase of less than 10% of the best known solutions while increasing the robustness in at least 20% for practically every benchmark set. And, in an attempt to solve larger instances with 50 or 100 activities, we also used the MRCPSP/max benchmark sets, where the increase of the makespan is approximately 35% with respect to the best known solutions at the same time as with a 20% increase in robustness.

Portfolio optimization using improved artificial bee colony approach

Nature-inspired optimization methods have been known to have the capability of handling computationally complicated problems, especially when traditional methods have become insufficient to. In this work, we proposed an improved artificial bee colony (IABC) method as the solution approach to trace out an efficiency frontier of the general portfolio performance. Such portfolio optimization problem focuses on balancing the trade-off between risk and return and is also captured in multidimensional nature with cardinality and bounding constraints. The proposed IABC algorithm intends to balance the diversity and quality of solutions, and fulfill the characteristic of the portfolio optimization problem. To do so, we employ a hybrid encoding that mixes integer and real variables in the IABC algorithm, and test its performance on four global stock market indices from the OR-Library. In addition, computational results are compared among four other algorithms. Evidences indicate that IABC performs the best in terms of diversity, convergence, and effectiveness among all four test data sets. The effect of choosing different number of stocks to form a portfolio is also investigated. The results confirm that less number of stocks selected in a portfolio can help to build a better efficiency frontier with lower risk and higher return more quickly.

A memetic algorithm with a variable block insertion heuristic for single machine total weighted tardiness problem with sequence dependent setup times

In this paper, a memetic algorithm with a variable block insertion heuristic is presented to solve the single machine total weighted tardiness problem with sequence dependent setup times. Together with the traditional insertion neighborhood structure, the memetic algorithm is combined with a variable block insertion heuristic in which a block of jobs are removed from a sequence and then inserted into all possible positions of the partial sequence. For this purpose, we devise a variable neighborhood descent algorithm to incorporate different block insertion heuristics having different block sizes. We also employ a simulated annealing type of acceptance criterion to diversify the population. To evaluate its performance, the memetic algorithm is tested on a set of benchmark instances from the literature. The analyses of experimental results have shown highly effective performance of the memetic algorithm against the best performing algorithms from the literature. The proposed memetic algorithm was able to find 98 out 120 optimal solutions within reasonable CPU times.

Biotech firm valuation in an emerging market – evidence from Taiwan

Purpose – This paper aims to explore the relationship of various financial and non‐financial factors to corporate value and how these factors can be used for the purpose of firm valuation. The focus is placed on a developing high‐tech industry.Design/methodology/approach – The authors collect and compare data from companies within the time window of 1997 through 2010. The techniques of stepwise regression and back‐propagation neural network (BPNN) are applied to analyze this data, where the variables of operating profit margin, ROE, ROA, net income ratio, Tobins Q and stock price are chosen to indicate firm value.Findings – Each firm value variable appears to have a different set of estimator variables consisting of financial and non‐financial factors. The estimator variable in the set that has a high influence relative to the others tends to be financial factor. However, certain non‐financial factors appear to be considered as an estimator variable for different firm value variables more often than fina...

Economic optimization of resource-constrained project scheduling: a two-phase metaheuristic approach

This paper deals with the problem of project scheduling subject to multiple execution modes with non-renewable resources, and a model that handles some of monetary issues in real world applications. The objective is to schedule the activities to maximize the expected net present value (NPV) of the project, taking into account the activity costs, the activity durations, and the cash flows generated by successfully completing an activity. Owing to the combinatorial nature of this problem, the current study develops a hybrid of branch-and-bound procedure and memetic algorithm to enhance both mode assignment and activity scheduling. Modifications for the makespan minimization problem have been made through a set of benchmark problem instances. Algorithmic performance is rated on the maximization of the project NPV and computational results show that the two-phase hybrid metaheuristic performs competitively for all instances of different problem sizes.

Production Scheduling Tools to Prevent and Repair Disruptions in MRCPSP

Companies invest countless hours in planning project execution because it is a crucial component for their growth. However, regardless of all the considerations taken in the planning stage, uncertainty inherent to project execution leads to schedule disruptions, and even renders projects unfeasible. There is a vast amount of studies for generating baseline (predictive) schedules, yet, the literature regarding reactive scheduling for the Multi-Mode Resource Constrained Project Scheduling Problem (MRCPSP) is scant with only two previous studies found at the time of writing. In contrast, schedule disruption management has been thoroughly studied in the mass production environment, and regardless of the difficulties encountered, they will almost certainly be required to meet the levels planned. With this in mind, this study proposes an integrative (proactive and reactive) scheduling framework that uses the experience and methodologies developed in the production scheduling environment and apply it to the MRCPSP. The purpose of this framework is to be used on further empirical research.