• Communications for Statistical Applications and Methods
• /
• v.17 no.4
• /
• pp.527-540
• /
• 2010

Rule-based classification analysis is widely used for massive datamining because it is easy to understand and its algorithm is uncomplicated. In this classification analysis, majority vote of rules or weighted combination of rules using their supports are frequently used in order to combine rules. We propose a method to combine rules by using the multinomial distribution in this paper. Iterative proportional fitting algorithm is used to estimate the multinomial distribution which maximizes entropy constrained on rules' support. Simulation experiments show that this method can compete with other well known classification models in the case of two similar populations.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.19 no.2
• /
• pp.417-426
• /
• 2024

In general, the implementation of machine learning requires prior knowledge and experience with deep learning models, and substantial computational resources and time are necessary for data processing. As a result, machine learning encounters several limitations when deployed on embedded processors. To address these challenges, this paper introduces a novel approach where a genetic algorithm is applied to the convolution operation within the machine learning process, specifically for performing a selective convolution operation.In the selective convolution operation, the convolution is executed exclusively on pixels identified by a genetic algorithm. This method selects and computes pixels based on a ratio determined by the genetic algorithm, effectively reducing the computational workload by the specified ratio. The paper thoroughly explores the integration of genetic algorithms into machine learning computations, monitoring the fitness of each generation to ascertain if it reaches the target value. This approach is then compared with the computational requirements of existing methods.The learning process involves iteratively training generations to ensure that the fitness adequately converges.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.22 no.4
• /
• pp.361-379
• /
• 2011

This paper aims to propose the most appropriate method for calculating the journal PageRank in a domestic citation database. Korean journals show relatively high journal self-citation ratios and have many outgoing citations to external journals which are not included in the domestic citation database. Because the PageRank algorithm requires recursive calculation to converge, those two characteristics of domestic citation databases must be accounted for in order to measure the citation impact of Korean journals. Therefore, two PageRank calculation methods and four formulas for self-citation adjustment have been examined and tested for KSCD journals. The results of the correlation analysis and regression analysis show that the SCImago Journal Rank formula with the cr2 type self-citation adjustment method seems to be a more appropriate way to measure the relative impact of domestic journals in the Korean Science Citation Database.

• The Journal of the Korea Contents Association
• /
• v.10 no.7
• /
• pp.10-22
• /
• 2010

This study proposes a new method of multiple homographies estimation between two images. With a large proportion of outliers, RANSAC is a general and very successful robust parameter estimator. However it is limited by the assumption that a single model acounts for all of the data inliers. Therefore, it has been suggested to sequentially apply RANSAC to estimate multiple 2D projective transformations. In this case, because outliers stay in the correspondence data set through the estimation process sequentially, it tends to progress slowly for all models. And, it is difficult to parallelize the sequential process due to the estimation order by the number of inliers for each model. We introduce a guided sequential RANSAC algorithm, using the local model instances that have been obtained from RANSAC procedure, which is able to reduce the number of random samples and deal simultaneously with multiple models.

• Korean Journal of Remote Sensing
• /
• v.29 no.2
• /
• pp.173-182
• /
• 2013

The early Sea-viewing Wide Field-of-view Sensor(SeaWiFS) atmospheric correction algorithm which is the basis of the atmospheric correction algorithm for Geostationary Ocean Color Imager(GOCI) assumes that water-leaving radiances is negligible at near-infrared(NIR) wavelengths. For this reason, all of the satellite measured radiances at the NIR wavelengths are assigned to aerosol radiances. However that assumption would cause underestimation of water-leaving radiances if it were applied to turbid Case-2 waters. To overcome this problem, Management Unit of the North Sea Mathematical Models(MUMM) atmospheric correction algorithm has been developed for turbid waters. This MUMM algorithm introduces new parameter ${\alpha}$, representing the ratio of water-leaving reflectance at the NIR wavelengths. ${\alpha}$ is calculated by statistical method and is assumed to be constant throughout the study area. Using this algorithm, we can obtain comparatively accurate water-leaving radiances in the moderately turbid waters where the NIR water-leaving reflectance is less than approximately 0.01. However, this algorithm still underestimates the water-leaving radiances at the extremely turbid water since the ratio of water-leaving radiance at two NIR wavelengths, ${\alpha}$ is changed with concentration of suspended particles. In this study, we modified the MUMM algorithm to calculate appropriate value for ${\alpha}$ using an iterative technique. As a result, the accuracy of water-leaving reflectance has been significantly improved. Specifically, the results show that the Root Mean Square Error(RMSE) of the modified MUMM algorithm was 0.002 while that of the MUMM algorithm was 0.0048.