• Journal of the Korean Society for Precision Engineering
• /
• v.16 no.1 s.94
• /
• pp.49-60
• /
• 1999

This paper addresses a welding task sequencing for robot arc welding process planning. Although welding task sequencing is an essential step in the welding process planning, it has not been considered through a systematic approach, but it depends rather on empirical knowledge. Thus, an effective task sequencing for robot arc welding is required. Welding perations can be classified by the number of welding robots. Genetic algorithms are applied to tackle those welding task sequencing problems. A genetic algorithm for traveling salesman problem (TSP) is utilized to determine welding task sequencing for a MultiWeldline-SingleLayer problem. Further, welding task sequencing for multiWeldline-MultiLayer welding is investigated and appropriate genetic algorithms are introduced. A random key genetic algorithm is also proposed to solve multi-robot welding sequencing : MultiWeldline with multi robots. Finally, the genetic algorithm are implemented for the welding task sequencing of three dimensional weld plate assemblies. Robot welding operations conforming to the algorithms are simulated in graphic detail using a robot simulation software IGRIP.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.4 no.1
• /
• pp.69-77
• /
• 1979

A single machine stochastic scheduling problem is considered. Associated with each job is its random processing time and random deferral cost. The criterion is to order the jobs so as to minimize the sum of the deferral costs. The expected sum of the deferral costs is theroretically derived under the stochastic situation for each of several scheduling decision rules which are well known for the deterministic environment. It is also shown that certain stochastic problems can be reduced to equivalent deterministic problems. Two examples are illustrated to show the expected total deferral costs.

• Journal of Microbiology and Biotechnology
• /
• v.20 no.12
• /
• pp.1696-1701
• /
• 2010

The ${\alpha}$-amylases activity was improved by random mutagenesis and screening. A region comprising residues from the position 34-281 was randomly mutated in B. licheniformis ${\alpha}$-amylase (AmyL), and the library with mutations ranging from low, medium, and high frequencies was generated. The library was screened using an effective liquid-phase screening method to isolate mutants with an altered pH profile. The sequencing of improved variants indicated 2-5 amino acid changes. Among them, mutant TP8H5 showed an altered pH profile as compared with that of wild type. The sequencing of variant TP8H5 indicated 2 amino acid changes, Ile157Ser and Trp193Arg, which were located in the solvent accessible flexible loop region in domain B.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2000.10a
• /
• pp.207-210
• /
• 2000

지금까지 인간이나 다른 생물체의 전체 유전체 염기서열을 밝혀내는 작업은 크게 세가지 방법으로 진행되었다. Clone-by-clone approach, sequence tagged connector approach, random shotgun approach(1)가 그것인데 마지막의 random shotgun approach는 fragment assembly problem을 비롯한 여러 가지 전산학적인 문제들을 수반한다. 이 논문은 저자들의 국내 최초로 미생물체의 전체 염기서열을 random shotgun approach를 이용하여 밝혀낸 경험을 바탕으로 그에 따르는 문제인 fragment assembly problem에 대해 소개하고 그에 수반되는 몇 가지 전산학적인 문제와 몇 가지 해결책에 대해 설명하려 한다.

• Journal of Korea Artificial Intelligence Association
• /
• v.2 no.1
• /
• pp.7-14
• /
• 2024

In this paper, we explore the application of RNA-sequencing data and ensemble machine learning to predict lung cancer and treatment strategies for lung cancer, a leading cause of cancer mortality worldwide. The research utilizes Random Forest, XGBoost, and LightGBM models to analyze gene expression profiles from extensive datasets, aiming to enhance predictive accuracy for lung cancer prognosis. The methodology focuses on preprocessing RNA-seq data to standardize expression levels across samples and applying ensemble algorithms to maximize prediction stability and reduce model overfitting. Key findings indicate that ensemble models, especially XGBoost, substantially outperform traditional predictive models. Significant genetic markers such as ADGRF5 is identified as crucial for predicting lung cancer outcomes. In conclusion, ensemble learning using RNA-seq data proves highly effective in predicting lung cancer, suggesting a potential shift towards more precise and personalized treatment approaches. The results advocate for further integration of molecular and clinical data to refine diagnostic models and improve clinical outcomes, underscoring the critical role of advanced molecular diagnostics in enhancing patient survival rates and quality of life. This study lays the groundwork for future research in the application of RNA-sequencing data and ensemble machine learning techniques in clinical settings.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.17
• /
• pp.7935-7941
• /
• 2015

The aim of this study was to implement massive parallel sequencing (MPS) technology in clinical genetics testing. We developed and tested an amplicon-based method for resequencing the BRCA1 and BRCA2 genes on an Illumina MiSeq to identify disease-causing mutations in patients with hereditary breast or ovarian cancer (HBOC). The coding regions of BRCA1 and BRCA2 were resequenced in 96 HBOC patient DNA samples obtained from different sample types: peripheral blood leukocytes, whole blood drops dried on paper, and buccal wash epithelia. A total of 16 random DNA samples were characterized using standard Sanger sequencing and applied to optimize the variant calling process and evaluate the accuracy of the MPS-method. The best bioinformatics workflow included the filtration of variants using GATK with the following cut-offs: variant frequency >14%, coverage ($>25{\times}$) and presence in both the forward and reverse reads. The MPS method had 100% sensitivity and 94.4% specificity. Similar accuracy levels were achieved for DNA obtained from the different sample types. The workflow presented herein requires low amounts of DNA samples (170 ng) and is cost-effective due to the elimination of DNA and PCR product normalization steps.

• Plant Breeding and Biotechnology
• /
• v.6 no.4
• /
• pp.426-433
• /
• 2018

Plant tissue culture is a technique that has invariably been used for various purposes such as obtaining transgenic plants for crop improvement or functional analysis of genes. However, this process can be associated with a variety of genetic and epigenetic instabilities in regenerated plants, termed as somaclonal variation. In this study, we investigated mutation spectrum, chromosomal distributions of nucleotide substitution types of single-nucleotide polymorphisms (SNPs) and insertions/deletions (InDels) by whole genome re-sequencing between Dongjin and Nipponbare along with regenerated plants of Dongjin from different induction periods. Results indicated that molecular spectrum of mutations in regenerated rice against Dongjin genome ranged from $9.14{\times}10^{-5}$ to $1.37{\times}10^{-4}$ during one- to three-month callus inductions, while natural mutation rate between Dongjin and Nipponbare genomes was $6.97{\times}10^{-4}$. Non-random chromosome distribution of SNP and InDel was observed in both regenerants and Dongjin genomes, with the highest densities on chromosome 11. The transition to transversion ratio was 2.25 in common SNPs of regenerants against Dongjin genome with the highest C/T transition frequency, which was similar to that of Dongjin against Nipponbare genome.

• Journal of Korean Institute of Industrial Engineers
• /
• v.9 no.1
• /
• pp.23-31
• /
• 1983

A group scheduling problem, which is production scheduling problem associated with the concept of group technology, is studied under due date constraints in multi-stage manufacturing systems. The purpose of this paper is to develop and evaluate a practical heuristic procedure for determining group sequence and job sequence within each group to minimize total tardiness in multi-stage manufacturing systems. A heuristic algorithm has been developed by introducing the concept of relative measures of job tardiness and group tardiness for job sequencing and group sequencing, respectively. A numerical example is shown to illustrate the proposed procedure. The heuristic algoirthm is tested by comparisons with problems with known optimal solutions and also with random group schedules for a set of large-size problems. Results indicate that the proposed heuristic algorithm provides good solutions with small computational requirements, and thus is viable for large size problems in practice.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.19 no.37
• /
• pp.201-210
• /
• 1996

Although FMS implementation in Korea is not yet mature, the worldwide empirical data shows the diffusion of FMS is inevitable in near future. As the reletionships between the high capital cost and the relative benefits and advantages are complex to analyse, it is rather beneficial to prepare the effective operation strategies which exploit the FMS flexibility, such as machine loading with alternative routing and dispatching rules. This paper shows the formulation applying a goal programming model for the loading problem with objectives of minimizing the production cost and maximizing the machine utilization, including constraints such as machine tool capacity and demands, etc. A realistic random FMS model is developed for illustration. Since loading and dispatching are a composite of two interdependent tasks, simulation is made to investigate the interactions between the two.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.23 no.3
• /
• pp.91-108
• /
• 1998

Automated Storage and Retrieval Systems (AS/RSs) are an important facility for modern material management. The expected benefits of these capital-intensive facilities are gained when their control policies and their physical design parameters are determined simultaneously. In this paper we present several analytical models that capture the impact of the storage assignment policy and of the rack design on the expected storage and retrieval times. Sequential and interleaved service modes are considered for sequencing the storage and retrieval requests. We further investigate the impact of the rack structure on the relative performance of the following storage assignment policies : closest open location (random), full turnover-based policy, and class-based. Our analysis clearly indicates that significant savings in crane travel time are realized when implementing full turnover-based policy, rather than random. These savings become more and more pronounced as the profile of the storage racks approaches the square-in-time shape. Furthermore, it is shown that a class-based policy, with a small number of storage classes, will capture most of these savings and be easier to manage in practice.