• Journal of the Korea Society of Computer and Information
• /
• v.21 no.11
• /
• pp.121-126
• /
• 2016

In recommender systems based on collaborative filtering, measuring similarity is very critical for determining the range of recommenders. Data sparsity problem is fundamental in collaborative filtering systems, which is partly solved by Jaccard coefficient combined with traditional similarity measures. This study proposes a new coefficient for improving performance of Jaccard coefficient by compensating for its drawbacks. We conducted experiments using datasets of various characteristics for performance analysis. As a result of comparison between the proposed and the similarity metric of Pearson correlation widely used up to date, it is found that the two metrics yielded competitive performance on a dense dataset while the proposed showed much better performance on a sparser dataset. Also, the result of comparing the proposed with Jaccard coefficient showed that the proposed yielded far better performance as the dataset is denser. Overall, the proposed coefficient demonstrated the best prediction and recommendation performance among the experimented metrics.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2000.07a
• /
• pp.950-952
• /
• 2000

Considering the effective distribution coefficient of Ni in Fe-Ni-Co invar alloy containing a little amount of carbon, we investigated on the thermal expansion coefficient(${\alpha}$). Fe-Ni-Co invar alloy had a large thermal expansion coefficient in as-casted compared with solution treated. The thermal expansion coefficient of Fe-Ni-Co alloy increased with the carbon content in both state of as-casted and solution treated. The effective distribution coefficient(Ke$\^$Ni/) of Ni was smaller than unity in alloy of not containing carbon, but is way larger than unity in alloy of containing carbon. It was considered that the homogeneity of Ni in primary austenite affected thermal expansion coefficient.

• Proceedings of the KSR Conference
• /
• 1998.11a
• /
• pp.512-521
• /
• 1998

The roll characteristic of railway vehicle is an important factor that affects the roll-over of vehicle and lateral ride comfort of passenger. Generally the roll characteristics of railway vehicle is defined by the term of roll-coefficient, s, which represents the ratio of incline or carbody to that of rail-cant. The limit values of roll coefficient recommended in UIC Bre 0.4 for coach without pantograph and 0.15 for vehicle with pantograph. The roll coefficient can be calculated by VAMPIRE that is the well-known commercial software for analysis of dynamic behavior of railway vehicle. The value of roll coefficient is effected by height of gravity center of carbody, stiffness of primary and secondary suspension and etc. The calculated roll-coefficient for electric locomotive and passenger coach is 0.12 and 0.77 respectively, The additional equipment such as anti-roll bar is considered in order to decrease roll-coefficient of passenger coach. In relation to roll characteristics, the analysis for roll-over due to wind is a1so performed. The results show that roll-characteristics affect the roll-over of vehicle.

• Journal of the Korean Statistical Society
• /
• v.30 no.1
• /
• pp.99-113
• /
• 2001

This paper deals with the problem of estimating the means in two inverse Gaussian populations with equal but unknown coefficient of variation. The maximum likelihood estimators are derived by solving a cubic equation and their asymptotic variances are presented for comparative purpose. Monte-Carlo simulation is conducted to investigate the efficiency of the estimators relative to the sample means over a wide range of values for the sample size and the coefficient of variation. The effect on this efficiency under the departure from the assumption of common coefficient of variation is also studied.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 1998.04a
• /
• pp.643-649
• /
• 1998

This paper is to compare the absorption coefficient with normal incidence waves and it with diffuse incidence waves of glass wool which the most used absorption material and multi layer fiber. The absorption coefficient with normal incidence waves is calculated by surface impedance that measured by using an acoustic tube. Based on this data, the absorption. coefficient with diffuse incidence waves is predicted and is compared with measured the absorption coefficient in reverberation room.

• Bulletin of the Korean Mathematical Society
• /
• v.46 no.3
• /
• pp.401-408
• /
• 2009

The object of the present paper is to study certain curvature restriction on an LP-Sasakian manifold with a coefficient $\alpha$. Among others it is shown that if an LP-Sasakian manifold with a coefficient $\alpha$ is a manifold of constant curvature, then the manifold is the product manifold. Also it is proved that a 3-dimensional Ricci semisymmetric LP-Sasakian manifold with a constant coefficient $\alpha$ is a spaceform.

• Proceedings of the IEEK Conference
• /
• 2000.07b
• /
• pp.571-574
• /
• 2000

A complex coefficient filter obtained by directly exchanging several reactance elements included in a real coefficient, filter for imaginary valued resistors is described. By using the proposed method, four varieties of complex coefficient filter are obtained. The stability problem is described. Finally, the frequency responses of the proposed kiters are shown.

• Communications for Statistical Applications and Methods
• /
• v.3 no.2
• /
• pp.77-83
• /
• 1996

This paper is concerned with an estimation problem for the AR(1) process $Y_t, t=0, {\pm}1, {\cdots}$with time carying autoregressive coefficient, where coefficient itself is also stochastic process. Attention is directed to the problem of finding a consistent estimator of ${\Phi}$, the mean level of autoregressive coefficient. The asymptotic distribution of the resulting consistent estimator of ${\Phi}$, is them discussed. We do not assume any time series model for the time varying autoregressive coefficient.

• Communications of the Korean Mathematical Society
• /
• v.16 no.3
• /
• pp.509-518
• /
• 2001

The purpose of this paper is to propose a numerical algorithm for the problem of coefficient identification of the scalar wave equation based on a local observation on the boundary: Determine the unknown coefficient function with the knowledge of simultaneous Dirichlet and Neumann boundary values on a part of boundary. To find the unknown coefficient function, the unknown Neumann boundary value is also identified. We recast our inverse problem to variational problem. The gradient method is applied to find the minimizing functions. We confirm the effectiveness of our algorithm by numerical experiments.

• Communications for Statistical Applications and Methods
• /
• v.14 no.2
• /
• pp.301-307
• /
• 2007

A effective teaching method on correlation coefficient for elementary level statistics course is discussed in this article. The well known inequalities, such as Theorem 368 of Hardy et al. (1952), are used for the interpretation of concept of covariance. An Excel example is provided for the illustration of concept of correlation coefficient.