• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2002.11a
• /
• pp.491-501
• /
• 2002

Case-based reasoning is emerging as a leading methodology for the application of artificial intelligence. CBR is a reasoning methodology that exploits similar experienced solutions, in the form of past cases, to solve new problems. Hybrid model achieves some convergence of the wide proliferation of credit evaluation modeling. As a result, Hybrid model showed that proposed methodology classify more accurately than any of techniques individually do. It is confirmed that proposed methodology predicts significantly better than individual techniques and the other combining methodologies. The objective of the proposed approach is to determines a set of weighting values that can best formalize the match between the input case and the previously stored cases and integrates fuzzy sit concepts into the case indexing and retrieval process. The GA is used to search for the best set of weighting values that are able to promote the association consistency among the cases. The fitness value in this study is defined as the number of old cases whose solutions match the input cases solution. In order to obtain the fitness value, many procedures have to be executed beforehand. Also this study tries to transform financial values into category ones using fuzzy logic approach fur performance of credit evaluation. Fuzzy set theory allows numerical features to be converted into fuzzy terms to simplify the matching process, and allows greater flexibility in the retrieval of candidate cases. Our proposed model is to apply an intelligent system for bankruptcy prediction.

• Journal of Korea Multimedia Society
• /
• v.17 no.1
• /
• pp.26-33
• /
• 2014

Video retrieval and indexing are performed by comparing feature similarities between key frames in shot after detecting a scene change and extracting key frames from the shot. Typical image features such as color, shape, and texture are used in content-based video and image retrieval. Many approaches for integrating these features have been studied. However, the issue of these approaches is how to appropriately assign weighting of key frame features at query time. Therefore, we propose a new video retrieval method using adaptively weighted image features. We performed computer simulations in test databases which consist of various kinds of key frames. The experimental results show that the proposed method has better performance than previous works in respect to several performance evaluations such as precision vs. recall, retrieval efficiency, and ranking measure.

• Communications for Statistical Applications and Methods
• /
• v.16 no.1
• /
• pp.103-113
• /
• 2009

Weighting is a common form of unit nonresponse adjustment in sample surveys where entire questionnaires are missing due to noncontact or refusal to participate. A common approach computes the response weight as the inverse of the response rate within adjustment cells based on covariate information. In this paper, we consider the efficiency and robustness of nonresponse weight adjustment bated on the response propensity and predictive mean. In the simulation study based on 2000 Fishry Census in Korea, the root mean squared errors for assessing the various ways of forming nonresponse adjustment cell s are investigated. The simulation result suggest that the most important feature of variables for inclusion in weighting adjustment is that they are predictive of survey outcomes. Though useful, prediction of the propensity to response is a secondary. Also the result suggest that adjustment cells based on joint classification by the response propensity and predictor of the outcomes is productive.

• Journal of Korea Multimedia Society
• /
• v.5 no.5
• /
• pp.512-522
• /
• 2002

In this paper, we propose a hybrid method which is mixed with frame selection and weighting model rank method, based on GMM(gaussian mixture model), for real-time text-independent speaker recognition system. In the system, maximum likelihood estimation was used for GMM parameter optimization, and maximum likelihood was used for recognition basically Proposed hybrid method has two steps. First, likelihood score was calculated with speaker models and test data at frame level, and the difference is calculated between the biggest likelihood value and second. And then, the frame is selected if the difference is bigger than threshold. The second, instead of calculated likelihood, weighting value is used for calculating total score at each selected frame. Cepstrum coefficient and regressive coefficient were used as feature parameters, and the database for test and training consists of several data which are collected at different time, and data for experience are selected randomly In experiments, we applied each method to baseline system, and tested. In speaker recognition experiments, proposed hybrid method has an average of 4% higher recognition accuracy than frame selection method and 1% higher than W method, implying the effectiveness of it.

• Journal of KIISE:Software and Applications
• /
• v.31 no.8
• /
• pp.1092-1100
• /
• 2004

An E-mails have regarded as one of the most popular methods for exchanging information because of easy usage and low cost. Meanwhile, exponentially growing unwanted mails in user's mailbox have been raised as main problem. Recognizing this issue, Korean government established a law in order to prevent e-mail abuse. In this paper we suggest hybrid spam mail filtering system using weighted Bayesian classifier which is extended from naive Bayesian classifier by adding the concept of preprocessing and intelligent agents. This system can classify spam mails automatically by using training data without manual definition of message rules. Particularly, we improved filtering efficiency by imposing weight on some character by feature extraction from spam mails. Finally, we show efficiency comparison among four cases - naive Bayesian, weighting on e-mail header, weighting on HTML tags, weighting on hyperlinks and combining all of four cases. As compared with naive Bayesian classifier, the proposed system obtained 5.7% decreased precision, while the recall and F-measure of this system increased by 33.3% and 31.2%, respectively.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.6
• /
• pp.705-711
• /
• 2011

This paper proposes an effective algorithm of license plate recognition for industrial applications. By applying Canny edge detection on a vehicle image, it is possible to find a connected rectangular, which is a strong candidate for license plate. The color information of license plate separates plates into white and green. Then, OTSU binary image processing and foreground neighbor pixel propagation algorithm CLNF will be applied to each license plates to reduce noise except numbers and letters. Finally, through labeling, numbers and letters will be extracted from the license plate. Letter and number regions, separated from the plate, pass through mesh method and thinning process for extracting feature vectors by X-Y projection method. The extracted feature vectors are compared with the pre-learned weighting values by backpropagation neural network to execute final recognition process. The experiment results show that the proposed license plate recognition algorithm works effectively.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.12
• /
• pp.3383-3397
• /
• 2023

Scene graphs serve as semantic abstractions of images and play a crucial role in enhancing visual comprehension and reasoning. However, the performance of Scene Graph Generation is often compromised when working with biased data in real-world situations. While many existing systems focus on a single stage of learning for both feature extraction and classification, some employ Class-Balancing strategies, such as Re-weighting, Data Resampling, and Transfer Learning from head to tail. In this paper, we propose a novel approach that decouples the feature extraction and classification phases of the scene graph generation process. For feature extraction, we leverage a transformer-based architecture and design an adaptive calibration function specifically for predicate classification. This function enables us to dynamically adjust the classification scores for each predicate category. Additionally, we introduce a Distribution Alignment technique that effectively balances the class distribution after the feature extraction phase reaches a stable state, thereby facilitating the retraining of the classification head. Importantly, our Distribution Alignment strategy is model-independent and does not require additional supervision, making it applicable to a wide range of SGG models. Using the scene graph diagnostic toolkit on Visual Genome and several popular models, we achieved significant improvements over the previous state-of-the-art methods with our model. Compared to the TDE model, our model improved mR@100 by 70.5% for PredCls, by 84.0% for SGCls, and by 97.6% for SGDet tasks.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.6
• /
• pp.2996-3011
• /
• 2017

With rapid growth of web technology and dissemination of smart devices, social networking service(SNS) is widely used. As a result, huge amount of data are generated from SNS such as Twitter, and sentiment analysis of SNS data is very important for various applications and services. In the existing sentiment analysis based on the $Na{\ddot{i}}ve$ Bayes algorithm, a same number of attributes is usually employed to estimate the weight of each class. Moreover, uncountable and meaningless attributes are included. This results in decreased accuracy of sentiment analysis. In this paper two methods are proposed to resolve these issues, which reflect the difference of the number of positive words and negative words in calculating the weights, and eliminate insignificant words in the feature selection step using Multinomial $Na{\ddot{i}}ve$ Bayes(MNB) algorithm. Performance comparison demonstrates that the proposed scheme significantly increases the accuracy compared to the existing Multivariate Bernoulli $Na{\ddot{i}}ve$ Bayes(BNB) algorithm and MNB scheme.

• Journal of Information Technology Application
• /
• v.1
• /
• pp.173-211
• /
• 1999

The task of classification permeates all walks of life, from business and economics to science and public policy. In this context, nonlinear techniques from artificial intelligence have often proven to be more effective than the methods of classical statistics. The objective of knowledge discovery and data mining is to support decision making through the effective use of information. The automated approach to knowledge discovery is especially useful when dealing with large data sets or complex relationships. For many applications, automated software may find subtle patterns which escape the notice of manual analysis, or whose complexity exceeds the cognitive capabilities of humans. This paper explores the utility of a collaborative learning approach involving integrated models in the preprocessing and postprocessing stages. For instance, a genetic algorithm effects feature-weight optimization in a preprocessing module. Moreover, an inductive tree, artificial neural network (ANN), and k-nearest neighbor (kNN) techniques serve as postprocessing modules. More specifically, the postprocessors act as second0order classifiers which determine the best first-order classifier on a case-by-case basis. In addition to the second-order models, a voting scheme is investigated as a simple, but efficient, postprocessing model. The first-order models consist of statistical and machine learning models such as logistic regression (logit), multivariate discriminant analysis (MDA), ANN, and kNN. The genetic algorithm, inductive decision tree, and voting scheme act as kernel modules for collaborative learning. These ideas are explored against the background of a practical application relating to financial fraud management which exemplifies a binary classification problem.

• The Bulletin of The Korean Astronomical Society
• /
• v.35 no.2
• /
• pp.82.1-82.1
• /
• 2010

We have investigated the spatial configuration of stars within the tidal radius of metal poor globular cluster NGC 6626 in the bulge direction. Data were obtained in near-IR J,H,Ks bands with wide-field ($20'\times20'$) detector, WIRCam at CFHT. To trace the stellar density around target cluster, we sorted cluster's member stars by using a mask filtering algorithm and weighting the stars on the color-magnitude diagram. From the weighted surface density map, we found that the stellar spatial distributions within the tidal radius appear asymmetric and distorted features. Especially, we found that more prominent over-density features are extending toward the direction of Galactic plane rather than toward the directions of the Galactic center and its orbital motion. This orientation of the stellar density distribution can be interpreted with result of disk-shock effect of the Galaxy that the cluster had been experienced. Indeed, this over-density feature are well represented in the radial surface density profile for different angular sections. As one of the metal poor globular clusters with extended horizontal branch (EHB) in the bulge direction, NGC 6626 is kinematically decoupled from the normal clusters and known to have disk motion of peculiar motion. Thus, our result will be able to add further constraints to understand the origin of this cluster and the formation of bulge region in early universe.