Masaharu Munetomo

An adaptive parameter binary-real coded genetic algorithm for constraint optimization problems: Performance analysis and estimation of optimal control parameters

Real parameter constrained problems are an important class of optimization problems that are encountered frequently in a variety of real world problems. On one hand, Genetic Algorithms (GAs) are an efficient search metaheuristic and a prominent member within the family of Evolutionary Algorithms (EAs), which have been applied successfully to global optimization problems. However, genetic operators in their standard forms are blind to the presence of constraints. Thus, the extension of GAs to constrained optimization problems by incorporating suitable handing techniques is an active direction within GAs research. Recently, we have proposed a Binary Real coded Genetic Algorithm (BRGA). BRGA is a new hybrid approach that combines cooperative Binary coded GA (BGA) with Real coded GA (RGA). It employs an adaptive parameter-based hybrid scheme that distributes the computational power and regulates the interactions between the cooperative versions, which operate in a sequential time-interleaving manner. In this study, we aim to extend BRGA to constrained problems by introducing a modified dynamic penalty function into the architecture of BRGA. We use the CEC2010 benchmark suite of 18 functions to analyze the quality, time and scalability performance of BRGA. To investigate the effectiveness of the proposed modification, we compare the performance of BRGA under both the original and the modified penalty functions. Moreover, to demonstrate the performance of BRGA, we compare it with the performance of some other EAs from the literature. We also implement a robust parameter tuning procedure that relies on techniques from statistical testing, experimental design and Response Surface Methodology (RSM) to estimate the optimal values for the control parameters to secure a good performance by BRGA against specific problems at hand.

Advanced genetic algorithm to solve MINLP problems over GPU

In this paper we propose a many-core implementation of evolutionary computation for GPGPU (General-Purpose Graphic Processing Unit) to solve non-convex Mixed Integer Non-Linear Programming (MINLP) and non-convex Non Linear Programming (NLP) problems using a stochastic algorithm. Stochastic algorithms being random in their behavior are difficult to implement over GPU like architectures. In this paper we not only succeed in implementation of a stochastic algorithm over GPU but show considerable speedups over CPU implementations. The stochastic algorithm considered for this paper is an adaptive resolution approach to genetic algorithm (arGA), developed by the authors of this paper. The technique uses the entropy measure of each variable to adjust the intensity of the genetic search around promising individuals. Performance is further improved by hybridization with adaptive resolution local search (arLS) operator. In this paper, we describe the challenges and design choices involved in parallelization of this algorithm to solve complex MINLPs over a commodity GPU using Compute Unified Device Architecture (CUDA) programming model. Results section shows several numerical tests and performance measurements obtained by running the algorithm over an nVidia Fermi GPU. We show that for difficult problems we can obtain a speedup of up to 20× with double precision and up to 42× with single precision.

Parallelization strategies for evolutionary algorithms for MINLP

Two different parallelization strategies for evolutionary algorithms for mixed integer nonlinear programming (MINLP) are discussed and numerically compared in this contribution. The first strategy is to parallelize some internal parts of the evolutionary algorithm. The second strategy is to parallelize the MINLP function calls outside and independently of the evolutionary algorithm. The first strategy is represented here by a genetic algorithm (arGA) for numerical testing. The second strategy is represented by an ant colony optimization algorithm (MIDACO) for numerical testing. It can be shown that the first parallelization strategy represented by arGA is inferior to the serial version of MIDACO, even though if massive parallelization via GPGPU is used. In contrast to this, theoretical and practical tests demonstrate that the parallelization strategy of MIDACO is promising for cpu-time expensive MINLP problems, which often arise in real world applications.

Screening for FtsZ Dimerization Inhibitors Using Fluorescence Cross-Correlation Spectroscopy and Surface Resonance Plasmon Analysis

FtsZ is an attractive target for antibiotic research because it is an essential bacterial cell division protein that polymerizes in a GTP-dependent manner. To find the seed chemical structure, we established a high-throughput, quantitative screening method combining fluorescence cross-correlation spectroscopy (FCCS) and surface plasmon resonance (SPR). As a new concept for the application of FCCS to polymerization-prone protein, Staphylococcus aureus FtsZ was fragmented into the N-terminal and C-terminal, which were fused with GFP and mCherry (red fluorescent protein), respectively. By this fragmentation, the GTP-dependent head-to-tail dimerization of each fluorescent labeled fragment of FtsZ could be observed, and the inhibitory processes of chemicals could be monitored by FCCS. In the first round of screening by FCCS, 28 candidates were quantitatively and statistically selected from 495 chemicals determined by in silico screening. Subsequently, in the second round of screening by FCCS, 71 candidates were also chosen from 888 chemicals selected via an in silico structural similarity search of the chemicals screened in the first round of screening. Moreover, the dissociation constants between the highest inhibitory chemicals and Staphylococcus aureus FtsZ were determined by SPR. Finally, by measuring the minimum inhibitory concentration, it was confirmed that the screened chemical had antibacterial activity against Staphylococcus aureus, including methicillin-resistant Staphylococcus aureus (MRSA).

Toward a Genetic Algorithm Based Flexible Approach for the Management of Virtualized Application Environments in Cloud Platforms

Resource management in cloud platforms becomes an increasingly complex and daunting task surrounded by various challenges of stringent QoS requirements, service availability guaranteeing and escalating overhead of the infrastructure that resulted from operation costs and ecological impact. On the other hand, virtualization adds a greater flexibility to the resource managers in addressing such challenges. However, at the same time, it imposes a further challenge of added management complexity. Recently, we have proposed a resource management model for cloud platforms, which utilizes a new resource mapping formulation and relays on a hybrid virtualization framework in an attempt to realize a resource manager that intelligently adapts the available cloud resources to satisfy the conflicting objectives of the running applications and underlying infrastructures requirements. Moreover, we have proposed state of the art Binary-Real coded Genetic Algorithm (BRGA), which has been applied successfully to a wide spectrum of global and constrained optimization problems from the known benchmark suites. In this paper, we aim to proceed by proposing a mathematical model and a modified version of BRGA to validate our model. In addition, we aim to evaluate the feasibility, effectiveness and scalability of our approach through simulation experiments.

An improved binary-real coded genetic algorithm for real parameter optimization

Genetic algorithms (GAs) are vital members within the family biologically inspired algorithms. It has been proven that the performance of GAs is largely affected by the type of encoding schemes used to encode optimization problems. Binary and real encoding schemes are the most popular ones. However, it is still controversial to decide the superiority of one of them for GAs performance. Therefore, we have recently proposed binary-real coded GA (BRGA) that has the ability to use both encoding schemes at the same time. BRGA relays on a parameterized hybrid scheme to share the computational power and coordinate the cooperation between binary coded GA (BGA) and real coded GA (RGA). In this paper, we aim to evaluate the performance of BRGA systematically by utilizing CEC2005 benchmark of 25 problems and adopting a robust experimental analysis approach. The quality and time performance of BRGA against the benchmark suite and in comparison with original component algorithms (BGA and RGA) is reported discussed and analyzed. Moreover, the performance of BRGA is compared with other Evolutionary Algorithms (EAs) from the literature.

An adaptive resolution hybrid binary-real coded genetic algorithm

In genetic algorithms (GAs), is it better to use binary encoding schemes or floating point encoding schemes? In this article, we try to tackle this controversial question by proposing a novel algorithm that divides the computational power between two cooperative versions of GAs. These are a binary-coded GA (bGA) and a real-coded GA (rGA). The evolutionary search is primarily led by the bGA, which identifies promising regions in the search space, while the rGA increases the quality of the solutions obtained by conducting an exhaustive search throughout these regions. The resolution factor (R), which has a value that is increasingly adapted during the search, controls the interactions between the two versions. We conducted comparison experiments employing a typical benchmark function to prove the feasibility of the algorithm under the critical scenarios of increasing problem dimensions and decreasing precision power.

Towards Thought Control of Next-Generation Wearable Computing Devices

A new wearable computing era featuring devices such as Google Glass, smartwatches, and digital contact lenses is almost upon us, bringing with it usability issues that conventional human computer interaction (HCI) modalities cannot resolve. Brain computer interface (BCI) technology is also rapidly advancing and is now at a point where noninvasive BCIs are being used in games and in healthcare. Thought control of wearable devices is an intriguing vision and would facilitate more intuitive HCI; however, to achieve even a modicum of control BCI currently requires massive processing power that is not available on mobile devices. Cloud computing is a maturing paradigm in which elastic computing power is provided on demand over networks. In this paper, we review the three technologies and take a look at possible ways cloud computing can be harnessed to provide the computational power needed to facilitate practical thought control of next-generation wearable computing devices.

Multi-Level Autonomic Architecture for the Management of Virtualized Application Environments in Cloud Platforms

resource management in cloud platforms becomes an increasingly complex and daunting task surrounded by various challenges of stringent QoS requirements, service availability guaranteeing and escalating overhead of the infrastructure that resulted from operation costs and ecological effects. Virtualization adds a greater flexibility to the resource manager in addressing such challenges. However, it imposes a further challenge of added management complexity. So, in this brief paper, we attempt to address still an open question of how to employ virtualization techniques effectively to realize a resource manager that intelligently adapts cloud platforms resource usage to satisfy the conflicting objectives of running applications and underlying cloud infrastructures by proposing a novel multi-level architecture which relays on a hybrid virtualization framework. We describe its functional components and dataflow and highlight the next steps that we will adopt in order to realize it and evaluate its feasibility and effectiveness.

A Bayesian Optimization Algorithm for De Novo ligand design based docking running over GPU

A principal fragment-based design approach is De Novo ligand design at which small-molecule structures from a database of existing compounds (or compounds that could be made) are docked into the protein binding site following a virtual synthesis scheme. New virtual structures can easily be constructed from combinatorial building blocks. Typically, tens of thousands of orientations are generated for each ligand candidate, therefore global optimization algorithms are usually employed to search the chemical space by generating new molecular structures through probing many different fragments in a combinatorial fashion. We propose using Bayesian Optimization Algorithm (BOA), a meta-heuristic algorithm, in searching the combination of pre-docked fragments through minimizing the energy of ligand-receptor docking. We further introduce the use of GPU (Graphical Processing Unit) to overcome the very long time required in evaluating each possible fragment combination. We show how the GPU utilization enables experimenting larger fragments and target receptors for more complex instances. The experiments resulted in regenerating three drug-like compounds defined in the ZINC database as well as finding a new compound. The Results show how the nVidias Tesla C1060 GPU was utilized to accelerate the docking process by two orders of magnitude.