• Kyungpook Mathematical Journal
• /
• v.15 no.1
• /
• pp.33-36
• /
• 1975

• Kyungpook Mathematical Journal
• /
• v.11 no.1
• /
• pp.25-28
• /
• 1971

• Kyungpook Mathematical Journal
• /
• v.22 no.1
• /
• pp.109-112
• /
• 1982

• Korean Journal of Cognitive Science
• /
• v.28 no.4
• /
• pp.299-314
• /
• 2017

Predicting compound-protein interactions in-silico is significant for the drug discovery. In this paper, we propose an scalable machine learning model to predict compound-protein interaction. The key idea of this scalable machine learning model is the architecture of pairwise neural network model and feature embedding method from the raw data, especially for protein. This method automatically extracts the features without additional knowledge of compound and protein. Also, the pairwise architecture elevate the expressiveness and compact dimension of feature by preventing biased learning from occurring due to the dimension and type of features. Through the 5-fold cross validation results on large scale database show that pairwise neural network improves the performance of predicting compound-protein interaction compared to previous prediction models.

• The KIPS Transactions:PartC
• /
• v.13C no.2 s.105
• /
• pp.195-202
• /
• 2006

For applying sensor networking technology for our daily life, security service is essential, and pairwise key establishment is the key point for security. In this paper, we propose fairwise key establishment mechanism for secure coumunication in sensor networks. In the mechanism, we cluster the network field before deployment and predistribute key materials to normal sensor nodes and clusterheads. For clusterheads, more key materials are predistributed, and after deployment, sensor nodes which need to establish pairwise keys with other sensor nodes in different clusters make request for related key materials to their own clusterheads. Our proposal reduces the memory requirements for normal sensor nodes by distributing more information to clusterheads, and it raises the security level and resilience against node captures. In addition, it guarantees perfect pairwise key establishments for every pair of neighboring nodes and provides efficient and secure sensor communications.

• The Korean Journal of Applied Statistics
• /
• v.35 no.2
• /
• pp.265-283
• /
• 2022

MovieLens data consists of recorded movie evaluations that was often used to measure the evaluation score in the recommendation system research field. In this paper, we provide additional information obtained by clustering user-specific genre preference information through movie evaluation data and movie genre data. Because the number of movie ratings per user is very low compared to the total number of movies, the missing rate in this data is very high. For this reason, there are limitations in applying the existing clustering methods. In this paper, we propose a convex clustering-based method using the pairwise fused penalty motivated by the analysis of MovieLens data. In particular, the proposed clustering method execute missing imputation, and at the same time uses movie evaluation and genre weights for each movie to cluster genre preference information possessed by each individual. We compute the proposed optimization using alternating direction method of multipliers algorithm. It is shown that the proposed clustering method is less sensitive to noise and outliers than the existing method through simulation and MovieLens data application.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2004.11a
• /
• pp.158-166
• /
• 2004

Predicting the destination of a protein in a cell gives valuable information for annotating the function of the protein. Recent technological breakthroughs have led us to develop more accurate methods for predicting the subcellular localization of proteins. The most important factor in determining the accuracy of these methods, is a way of extracting useful features from protein sequences. We propose a new method for extracting appropriate features only from the sequence data by computing pairwise sequence alignment scores. As a classifier, support vector machine (SVM) is used. The overall prediction accuracy evaluated by the jackknife validation technique reach 94.70% for the eukaryotic non-plant data set and 92.10% for the eukaryotic plant data set, which show the highest prediction accuracy among methods reported so far with such data sets. Our numerical experimental results confirm that our feature extraction method based on pairwise sequence alignment, is useful for this classification problem.

• Proceedings of the Korean Society of Soil and Groundwater Environment Conference
• /
• 2003.04a
• /
• pp.259-262
• /
• 2003

In the Republic of Korea, Ministry of Construction & Transportation and Korea Water Resources Corporation manage the national groundwater monitoring network at the 169 stations and will organize the supplementary groundwater monitoring network at the 10,000 stations by 2011 year. The method that organizes the monitoring network was developed using the Analytic Hierarchy Process with pairwise comparison. Several estimation factors for the estimating every district were selected to reflect each district conditions. Their weighting value was decided by pairwise comparison and questions to the experts about groundwater The optimal number of groundwater monitoring well was calculated through the developed method. To verify this method, groundwater was monitored in Jeonju city by way showing the example. The study area In Jeonju city needs 7 stations for the supplementary groundwater monitoring network. The results monitored in 7 stations inferred the groundwater level around the study area by Kriging. The mean of residual between inferred groundwater level value from Kriging and actual groundwater level is rather low. Furthermore, the mean and standard deviation of residual between inferred groundwater level change and actual groundwater change is much lower. The Fact that 7 monitoring stations are sufficient for observing the groundwater condition in the study area makes it possible for suggested monitoring number to be proper.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.31 no.6C
• /
• pp.575-581
• /
• 2006

In this paper, we first define one-dimensional component symbol error function (ODSEF) from the exact expression of the pairwise error probability of orthogonal space-time block codes (OSTBC). Using the ODSEF and the general bit error probability (BEP) expression for quadrature amplitude modulation (QAM) introduced by Cho and Yoon, the exact closed form expressions for the BEP of linear OSTBCs with QAM in slow-varying Rayleigh fading channel are derived.

• Journal of Information Processing Systems
• /
• v.13 no.5
• /
• pp.1089-1102
• /
• 2017

In software systems, it has been observed that a fault is often caused by an interaction between a small number of input parameters. Even for moderately sized software systems, exhaustive testing is practically impossible to achieve. This is either due to time or cost constraints. Combinatorial (t-way) testing provides a technique to select a subset of exhaustive test cases covering all of the t-way interactions, without much of a loss to the fault detection capability. In this paper, an approach is proposed to generate 2-way (pairwise) test sets using genetic algorithms. The performance of the algorithm is improved by creating an initial solution using the overlap coefficient (a similarity matrix). Two mutation strategies have also been modified to improve their efficiency. Furthermore, the mutation operator is improved by using a combination of three mutation strategies. A comparative survey of the techniques to generate t-way test sets using genetic algorithms was also conducted. It has been shown experimentally that the proposed approach generates faster results by achieving higher percentage coverage in a fewer number of generations. Additionally, the size of the mixed covering arrays was reduced in one of the six benchmark problems examined.