• Bulletin of the Korean Mathematical Society
• /
• v.33 no.3
• /
• pp.377-383
• /
• 1996

We consider in this paper the following nonlinear regression model $$ (1.1) y_t = f(x_t, \theta_o) + \in_t, t = 1, \ldots, n, $$ where $y_t$ is the tth response, $x_t$ is m-vector imput variable, $\theta_o$ is a p-vector of unknown parameter belong to a parameter space $\Theta, f:R^m \times \Theta \ to R^1$ is a nonlinear known function, and $\in_t$ are independent unobservable random errors with finite second moment.

• Journal of the Korean Data and Information Science Society
• /
• v.16 no.4
• /
• pp.1087-1094
• /
• 2005

e-insensitive support vector regression(e-SVR) is capable of providing more complete description of the linear and nonlinear relationships among random variables. In this paper we propose an iterative reweighted least squares(IRWLS) procedure to solve the quadratic problem of e-SVR with a modified loss function. Furthermore, we introduce the generalized approximate cross validation function to select the hyperparameters which affect the performance of e-SVR. Experimental results are then presented which illustrate the performance of the IRWLS procedure for e-SVR.

• Communications for Statistical Applications and Methods
• /
• v.22 no.2
• /
• pp.201-207
• /
• 2015

Support vector machine (SVM) is a well known kernel-trick supervised learning tool. This study proposes a working scheme for kernel-trick regression and classification (KtRC) as a SVM alternative. KtRC fits the model on a number of random subsamples and selects the best model. Empirical examples and a simulation study indicate that KtRC's performance is comparable to SVM.

• Journal of the Korean Mathematical Society
• /
• v.54 no.6
• /
• pp.1879-1903
• /
• 2017

In this paper we study the closure property and probability tail asymptotics for randomly weighted sums $S^{\Theta}_n={\Theta}_1X_1+{\cdots}+{\Theta}_nX_n$ for long-tailed random variables $X_1,{\ldots},X_n$ and positive bounded random weights ${\Theta}_1,{\ldots},{\Theta}_n$ under similar dependence structure as in [26]. In particular, we study the case where the distribution of random vector ($X_1,{\ldots},X_n$) is generated by an absolutely continuous copula.

• Industrial Engineering and Management Systems
• /
• v.15 no.4
• /
• pp.432-445
• /
• 2016

A random profile is the result of a process, the output of which is a function instead of a scalar or vector quantity. In the nature of these objects, two main dimensions of "functionality" and "randomness" can be recognized. Valuable researches have been conducted to present control charts for monitoring such processes in which a regression approach has been applied by focusing on "randomness" of profiles. Performing other statistical techniques such as hypothesis testing for different parameters, comparing parameters of two populations, ANOVA, DOE, etc. has been postponed thus far, because the "functional" nature of profiles is ignored. In this paper, first, some needed theorems are proven with an applied approach, so that be understandable for an engineer which is unfamiliar with advanced mathematical analysis. Then, as an application of that, a statistical test is designed for mean of continuous random profiles. Finally, using experimental operating characteristic curves obtained in computer simulation, it is demonstrated that the presented tests are properly able to recognize deviations in the null hypothesis.

• Communications for Statistical Applications and Methods
• /
• v.13 no.2
• /
• pp.243-254
• /
• 2006

The relative frequency of order statistics is investigated for independent and identically distributed (i.i.d.) random variables. Specifically, it is shown that the probability $Pr\{X_{[s]}=x\}$ is no less than the probability $Pr\{X_{[r]}=x\}$ at any point $x{\geqq}x_0$ when r$X_{[r]}$ denotes the r-th order statistic of an i.i.d. discrete random vector and $x_0$ depends on the population probability distribution. A similar result for i.i.d. continuous random vectors is also presented.

• The Pure and Applied Mathematics
• /
• v.11 no.2
• /
• pp.139-147
• /
• 2004

Let {<$\mathds{X}_t$} be an m-dimensional linear process of the form $\mathbb{X}_t\;=\sumA,\mathbb{Z}_{t-j}$ where {$\mathbb{Z}_t$} is a sequence of stationary m-dimensional negatively associated random vectors with $\mathbb{EZ}_t$ = $\mathbb{O}$ and $\mathbb{E}\parallel\mathbb{Z}_t\parallel^2$ < $\infty$. In this paper we prove the central limit theorems for multivariate linear processes generated by negatively associated random vectors.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.12
• /
• pp.6161-6176
• /
• 2018

Pseudorandom numbers are useful in cryptographic operations for using as nonce, initial vector, secret key, etc. Security of the cryptosystem relies on the secret key parameters, so a good pseudorandom number is needed. In this paper, we have proposed a new approach for generation of pseudorandom number. This method uses the three dimensional combinational puzzle Rubik Cube for generation of random numbers. The number of possible combinations of the cube approximates to 43 quintillion. The large possible combination of the cube increases the complexity of brute force attack on the generator. The generator uses cryptographic hash function. Chaotic map is being employed for increasing random behavior. The pseudorandom sequence generated can be used for cryptographic applications. The generated sequences are tested for randomness using NIST Statistical Test Suite and other testing methods. The result of the tests and analysis proves that the generated sequences are random.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.2
• /
• pp.391-399
• /
• 2013

Partially linear regression is capable of providing more complete description of the linear and nonlinear relationships among random variables. In support vector regression (SVR) the hyper-parameters are known to affect the performance of regression. In this paper we propose an iterative reweighted least squares (IRWLS) procedure to solve the quadratic problem of partially linear support vector regression with a modified loss function, which enables us to use the generalized approximate cross validation function to select the hyper-parameters. Experimental results are then presented which illustrate the performance of the partially linear SVR using IRWLS procedure.

• Communications for Statistical Applications and Methods
• /
• v.18 no.2
• /
• pp.165-170
• /
• 2011

Support vector quantile regression(SVQR) is capable of providing a good description of the linear and nonlinear relationships among random variables. In this paper we propose a sparse SVQR to overcome a limitation of SVQR, nonsparsity. The asymmetric e-insensitive loss function is used to efficiently provide sparsity. The experimental results are presented to illustrate the performance of the proposed method by comparing it with nonsparse SVQR.