• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.34S no.10
• /
• pp.98-103
• /
• 1997

Given the small bit allocation for motion information in very low bit-rate coding, motion estimation using the block matching algorithm(BMA) fails to maintain an acceptable level of prediction errors. The reson is that the motion model, or spatial transformation, assumed in block matching cannot approximate the motion in the real world precisely with a small number of parameters. In order to overcome the drawback of the conventional block matching algorithm, several triangle-based methods which utilize triangular patches insead of blocks have been proposed. To estimate the motions of image sequences, these methods usually have been based on the combination of optical flow equation, affine transform, and iteration. But the compuataional cost of these methods is expensive. This paper presents a fast motion estimation algorithm using a multiple linear regression model to solve the defects of the BMA and the triange-based methods. After describing the basic 2-D triangle-based method, the details of the proposed multiple linear regression model are presented along with the motion estimation results from one standard video sequence, representative of MPEG-4 class A data. The simulationresuls show that in the proposed method, the average PSNR is improved about 1.24 dB in comparison with the BMA method, and the computational cost is reduced about 25% in comparison with the 2-D triangle-based method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.9 no.2
• /
• pp.152-158
• /
• 2003

A machine-vision guided inspection system with go-no gauges for inspecting eggcrate assemblies in steam generators is considered. To locate the gauge at the right place, periodic corrective actions for its position and traveling direction are required. We present a machine vision algorithm for determining inserting position and traveling direction of the go-no gauge. The overall procedure of the algorithm is composed of camera calibration, eggcrate image preprocessing, grid-height adjustment, intersection point estimation between two intersecting grids, and adjustment of position and traveling direction of the gauge. The intersection point estimation is performed by using linear regression with a constraint. A test with a real eggcrate specimen shows the feasibility of the algorithm.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.2
• /
• pp.190-198
• /
• 2002

In general, we have used the Viterbi algorithm of Speech recognition for decoding. But a decoder in speaker verification has to recognize same word of every speaker differently. In this paper, we propose a noble decoding algorithm that could replace the typical Viterbi algorithm for the speaker verification system. We utilize for the proposed algorithm the speaker adaptation algorithms that transform feature vectors into the region of the client' characteristics in the speech recognition. There are many adaptation algorithms, but we take MLLR (Maximum Likelihood Linear Regression) and MAP (Maximum A-Posterior) adaptation algorithms for proposed algorithm. We could achieve improvement of performance about 30% of EER (Equal Error Rate) using proposed algorithm instead of the typical Viterbi algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.8
• /
• pp.3962-3980
• /
• 2019

To deal with single sample face recognition, this paper presents a patch based semi-supervised linear regression (PSLR) algorithm, which draws facial variation information from unlabeled samples. Each facial image is divided into overlapped patches, and a regression model with mapping matrix will be constructed on each patch. Then, we adjust these matrices by mapping unlabeled patches to $[1,1,{\cdots},1]^T$. The solutions of all the mapping matrices are integrated into an overall objective function, which uses ${\ell}_{2,1}$-norm minimization constraints to improve discrimination ability of mapping matrices and reduce the impact of noise. After mapping matrices are computed, we adopt majority-voting strategy to classify the probe samples. To further learn the discrimination information between probe samples and obtain more robust mapping matrices, we also propose a multistage PSLR (MPSLR) algorithm, which iteratively updates the training dataset by adding those reliably labeled probe samples into it. The effectiveness of our approaches is evaluated using three public facial databases. Experimental results prove that our approaches are robust to illumination, expression and occlusion.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.448-448
• /
• 2018

In this study, the regression models (Load ESTimator and eight-parameter model) were evaluated to estimate instantaneous pollutant loads under various criteria and optimization methods. As shown in the results, LOADEST commonly used in interpolating pollutant loads could not necessarily provide the best results with the automatic selected regression model. It is inferred that the various regression models in LOADEST need to be considered to find the best solution based on the characteristics of watersheds applied. The recently developed eight-parameter model integrated with Genetic Algorithm (GA) and Gradient Descent Method (GDM) were also compared with LOADEST indicating that the eight-parameter model performed better than LOADEST, but it showed different behaviors in calibration and validation. The eight-parameter model with GDM could reproduce the nitrogen loads properly outside of calibration period (validation). Furthermore, the accuracy and precision of model estimations were evaluated using various criteria (e.g., $R^2$ and gradient and constant of linear regression line). The results showed higher precisions with the $R^2$ values closed to 1.0 in LOADEST and better accuracy with the constants (in linear regression line) closed to 0.0 in the eight-parameter model with GDM. In hence, based on these finding we recommend that users need to evaluate the regression models under various criteria and calibration methods to provide the more accurate and precise results for pollutant load estimations.

• Journal of the Korean Data and Information Science Society
• /
• v.15 no.2
• /
• pp.449-455
• /
• 2004

Support vector machine (SVM) has been very successful in pattern recognition and function estimation problems for crisp data. This paper proposes a new method to evaluate interval regression models for crisp input-output data. The proposed method is based on quadratic loss SVM, which implements quadratic programming approach giving more diverse spread coefficients than a linear programming one. The proposed algorithm here is model-free method in the sense that we do not have to assume the underlying model function. Experimental result is then presented which indicate the performance of this algorithm.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.21 no.7
• /
• pp.61-67
• /
• 2007

The load forecasting has been an important part of power system Accordingly, it has been proposed various methods for the load forecasting. The load patterns of the special days is quite different than those of ordinary weekdays. It is difficult to accurately forecast the load of special days due to the insufficiency of the load patterns compared with ordinary weekdays, so we have proposed fuzzy least squares linear regression algorithm for the load forecasting. In this paper we proposed four models for fuzzy least squares linear regression. It is separated by coefficients of fuzzy least squares linear regression equation. we compared model of H1 with H4 and prove it H4 has accurately forecast better than H1.

• The Korean Journal of Applied Statistics
• /
• v.34 no.6
• /
• pp.937-944
• /
• 2021

The problem of selecting variables is addressed in partial linear regression. Model selection for partial linear models is not easy since it involves nonparametric estimation such as smoothing parameter selection and estimation for linear explanatory variables. In this work, several approaches for variable selection are proposed using a fast forward selection algorithm, least angle regression (LARS). The proposed procedures use t-test, all possible regressions comparisons or stepwise selection process with variables selected by LARS. An example based on real data and a simulation study on the performance of the suggested procedures are presented.

• Korean Journal of Artificial Intelligence
• /
• v.9 no.1
• /
• pp.21-27
• /
• 2021

In this Paper, Since the 1990s, Korea's credit card industry has steadily developed. As a result, various problems have arisen, such as careless customer information management and loans to low-credit customers. This, in turn, had a high delinquency rate across the card industry and a negative impact on the economy. Therefore, in this paper, based on Azure, we analyze and predict the delinquency and delinquency periods of credit loans according to gender, own car, property, number of children, education level, marital status, and employment status through linear regression analysis and enhanced decision tree algorithm. These predictions can consequently reduce the likelihood of reckless credit lending and issuance of credit cards, reducing the number of bad creditors and reducing the risk of banks. In addition, after classifying and dividing the customer base based on the predicted result, it can be used as a basis for reducing the risk of credit loans by developing a credit product suitable for each customer. The predicted result through Azure showed that when predicting with Linear Regression and Boosted Decision Tree algorithm, the Boosted Decision Tree algorithm made more accurate prediction. In addition, we intend to increase the accuracy of the analysis by assigning a number to each data in the future and predicting again.

• Korean Journal of Artificial Intelligence
• /
• v.10 no.1
• /
• pp.9-14
• /
• 2022

The demand for public bicycles operated by the Seoul Metropolitan Government is increasing every year. The size of the Seoul public bicycle project, which first started with about 5,600 units, increased to 3,7500 units as of September 2021, and the number of members is also increasing every year. However, as the size of the project grows, excessive budget spending and deficit problems are emerging for public bicycle projects, and new bicycles, rental office costs, and bicycle maintenance costs are blamed for the deficit. In this paper, the Azure Machine Learning Studio program and the Boosted Decision Tree Regression technique are used to predict the number of public bicycle rental over environmental factors and time. Predicted results it was confirmed that the demand for public bicycles was high in the season except for winter, and the demand for public bicycles was the highest at 6 p.m. In addition, in this paper compare four additional regression algorithms in addition to the Boosted Decision Tree Regression algorithm to measure algorithm performance. The results showed high accuracy in the order of the First Boosted Decision Tree Regression Algorithm (0.878802), second Decision Forest Regression (0.838232), third Poison Regression (0.62699), and fourth Linear Regression (0.618773). Based on these predictions, it is expected that more public bicycles will be placed at rental stations near public transportation to meet the growing demand for commuting hours and that more bicycles will be placed in rental stations in summer than winter and the life of bicycles can be extended in winter.