• Journal of Intelligence and Information Systems
• /
• v.9 no.3
• /
• pp.65-79
• /
• 2003

In general, the extraction and recognition of identifier is very hard work, because the scale or location of identifier is not fixed-form. And, because the provided image is contained by camera, it has some noises. In this paper, we propose methods for automatic detecting edge using canny edge mask. After detecting edges, we extract regions of identifier by detected edge information's. In regions of identifier, we extract each identifier using contour tracking algorithm. The self-generation supervised learning algorithm is proposed for recognizing them, which has the algorithm of combining the enhanced ART1 and the supervised teaming method. The proposed method has applied to the container images. The extraction rate of identifier obtained by using contour tracking algorithm showed better results than that from the histogram method. Furthermore, the recognition rate of the self-generation supervised teaming method based on enhanced ART1 was improved much more than that of the self-generation supervised learning method based conventional ART1.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2007.11a
• /
• pp.383-387
• /
• 2007

Traditional clustering is regarded as an unsupervised teaming to analyze unlabeled data. Semi-supervised clustering uses a small amount of labeled data to predict labels of unlabeled data as well as to improve clustering performance. Previous methods use constraints generated from available labeled data in clustering process. We propose a new constraint-based semi-supervised clustering method by reflecting initial predicted labels of unlabeled data. We evaluate and compare the performance of the proposed method in terms of classification errors through numerical experiments with blinded labeled data.

• Journal of KIISE:Software and Applications
• /
• v.35 no.5
• /
• pp.325-334
• /
• 2008

The goal of text categorization is to classify documents into a certain number of pre-defined categories. The previous studies in this area have used a large number of labeled training documents for supervised learning. One problem is that it is difficult to create the labeled training documents. While it is easy to collect the unlabeled documents, it is not so easy to manually categorize them for creating training documents. In this paper, we propose a new text categorization method based on semi-supervised learning. The proposed method uses only unlabeled documents and keywords of each category, and it automatically constructs training data from them. Then a text classifier learns with them and classifies text documents. The proposed method shows a similar degree of performance, compared with the traditional supervised teaming methods. Therefore, this method can be used in the areas where low-cost text categorization is needed. It can also be used for creating labeled training documents.

• Proceedings of the IEEK Conference
• /
• 2001.06c
• /
• pp.9-12
• /
• 2001

Learning vector quantization (LVQ) is a supervised learning technique that uses class information to move the vector quantizer slightly, so as to improve the quality of the classifier decision regions. In this paper we propose a selection method of initial codebook vectors for a teaming vector quantization (LVQ3) using support vector machines (SVM). The method is experimented with artificial and real design data sets and compared with conventional methods of the condensed nearest neighbor (CNN) and its modifications (mCNN). From the experiments, it is discovered that the proposed method produces higher performance than the conventional ones and then it could be used efficiently for designing nonparametric classifiers.

• Proceedings of the KIEE Conference
• /
• 1989.07a
• /
• pp.495-499
• /
• 1989

Concept is the set of selected neurons in a stable state of a neurel network. The Recognitron uses a parallel feedback structure to support concept learning. A number of clusters can exist in response to a given input, each of which make up a selective neuron. There are supervised and unsupervised learnig methods in concept teaming. In this paper, we have chosen unsupervised learning. Also, a new concept called relaxational learning has been introduced to stop runaway weights

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.8
• /
• pp.1743-1750
• /
• 2003

DC motor requires a rotor speed sensor for accurate speed control. The speed sensors such as resolvers and encoders are used as a speed detector, but they increase cost and size of the motor and restrict the industrial drive applications. So in these days, many papers have reported in the sensorless operation of DC motor〔3­5〕. This paper presents a new sensorless strategy using neural networks〔6­8〕. Neural network has three layers which are input layer, hidden layer and output layer. The optimal neural network structure was tracked down by trial and error, and it was found that 4­16­1 neural network structure has given suitable results for the instantaneous rotor speed. Also, learning method is very important in neural network. Supervised learning methods〔8〕 are typically used to train the neural network for learning the input/output pattern presented. The back­propagation technique adjusts the neural network weights during training. The rotor speed is gained by weights and four inputs to the neural network. The experimental results were found satisfactory in both the independency on machine parameters and the insensitivity to the load condition.

• Journal of KIISE:Software and Applications
• /
• v.31 no.7
• /
• pp.849-858
• /
• 2004

Most of supervised teaming algorithms could be applied after that continuous variables are transformed to categorical ones at the preprocessing stage in order to avoid the difficulty of processing continuous variables. This preprocessing stage is called global discretization, uses the class distribution list called bins. But, when data are large and the range of the variable to be discretized is very large, many sorting and merging should be performed to produce a single bin because most of global discretization methods need a single bin. Also, if new data are added, they have to perform discretization from scratch to construct categories influenced by the data because the existing methods perform discretization in batch mode. This paper proposes a method that extracts sample points and performs discretization from these sample points in order to solve these problems. Because the approach in this paper does not require merging for producing a single bin, it is efficient when large data are needed to be discretized. In this study, an experiment using real and synthetic datasets was made to compare the proposed method with an existing one.

• Journal of KIISE:Software and Applications
• /
• v.31 no.4
• /
• pp.447-456
• /
• 2004

Information extraction is a process of recognizing and fetching particular information fragments from a document. In order to extract information uniformly from many heterogeneous information sources, it is necessary to produce information extraction rules called a wrapper for each source. Previous methods of information extraction can be categorized into manual wrapper generation and automatic wrapper generation. In the manual method, since the wrapper is manually generated by a human expert who analyzes documents and writes rules, the precision of the wrapper is very high whereas it reveals problems in scalability and efficiency In the automatic method, the agent program analyzes a set of example documents and produces a wrapper through learning. Although it is very scalable, this method has difficulty in generating correct rules per se, and also the generated rules are sometimes unreliable. This paper tries to combine both manual and automatic methods by proposing a new method of learning information extraction rules. We adopt the scheme of supervised learning in which a user-interface agent is designed to get information from the user regarding what to extract from a document, and eventually XML-based information extraction rules are generated through learning according to these inputs. The interface agent is used not only to generate new extraction rules but also to modify and extend existing ones to enhance the precision and the recall measures of the extraction system. We have done a series of experiments to test the system, and the results are very promising. We hope that our system can be applied to practical systems such as information-mediator agents.

• Journal of KIISE:Software and Applications
• /
• v.32 no.8
• /
• pp.693-703
• /
• 2005

Given the local traders with pattern-based multi-predictors of stock prices, we study a method of dynamic asset allocation to maximize the trading performance. To optimize the proportion of asset allocated to each recommendation of the predictors, we design an asset allocation strategy called meta policy in the reinforcement teaming framework. We utilize both the information of each predictor's recommendations and the ratio of the stock fund over the total asset to efficiently describe the state space. The experimental results on Korean stock market show that the trading system with the proposed meta policy outperforms other systems with fixed asset allocation methods. This means that reinforcement learning can bring synergy effects to the decision making problem through exploiting supervised-learned predictors.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.3
• /
• pp.684-689
• /
• 2005

In this paper, we propose car license plate recognition using morphological information and an enhanced neural network. Morphological information on horizontal and vertical edges was used to extract the license plate from a car image. We used a contour tracking algorithm combined with the method of histogram and location information to extract individual characters in the extracted plate. The enhanced neural network is proposed for recognizing them, which has the method of combining the ART-1 and the supervised teaming method. The proposed method has applied to real world car images. The experimental results show that the proposed method has better the extraction rates than the methods with information of the thresholding, the RGB and the HSI, respectively. And the proposed neural network has better recognition performance than the conventional neural networks.