• Proceedings of the KIEE Conference
• /
• 1989.07a
• /
• pp.481-485
• /
• 1989

The object of this parer is on the design of an broad board classifier for connected. Korean digit. Many approaches have been applied in speech recognition systems: parametric vector quantization, dynamic programming and hiden Markov model. In the 80's the neural network method, which is expected to solve complex speech recognition problems, came bach. We have chosen the rule based system for our model. The phoneme-groups that we wish to classify are vowel_like, plosive_like fricative_like, and stop_like.The data used are 1380 connected digits spoken by three untrained male speakers. We have seen 91.5% classification rate.

• The Journal of the Korea Contents Association
• /
• v.10 no.5
• /
• pp.99-106
• /
• 2010

Multilayer perceptrons(MLPs) or feed-forward neural networks are widely applied to many areas based on their function approximation capabilities. When implementing MLPs for application problems, we should determine various parameters and training methods. In this paper, we discuss the design of MLPs especially for pattern classification problems. This discussion includes how to decide the number of nodes in each layer, how to initialize the weights of MLPs, how to train MLPs among various error functions, the imbalanced data problems, and deep architecture.

• Proceedings of the IEEK Conference
• /
• summer
• /
• pp.371-375
• /
• 2004

Fault diagnosis is one of the most studied problems in process engineering. Recently, great research interest has been devoted to approaches that use classification methods to detect faults. This paper presents an application of a newly developed classification method - support vector machines - for fault diagnosis in an industrial case. A real set of operation data of a motor pump was used to train and test the support vector machines. The experiment results show that the support vector machines give higher correct detection rate of faults in comparison to rule-based diagnostics. In addition, the studied method can work with fewer training instances, what is important for online diagnostics.

• Journal of Korea Multimedia Society
• /
• v.22 no.6
• /
• pp.665-675
• /
• 2019

In this paper, we compare and analyze the classification performance of deep learning algorithm Convolutional Neural Network(CNN) ac cording to ensemble generation and combining techniques. We used several CNN models(VGG16, VGG19, DenseNet121, DenseNet169, DenseNet201, ResNet18, ResNet34, ResNet50, ResNet101, ResNet152, GoogLeNet) to create 10 ensemble generation combinations and applied 6 combine techniques(average, weighted average, maximum, minimum, median, product) to the optimal combination. Experimental results, DenseNet169-VGG16-GoogLeNet combination in ensemble generation, and the product rule in ensemble combination showed the best performance. Based on this, it was concluded that ensemble in different models of high benchmarking scores is another way to get good results.

• Journal of Internet Computing and Services
• /
• v.8 no.3
• /
• pp.131-143
• /
• 2007

The current trends in information technology and intelligent systems use data mining techniques to discover patterns and extract rules from distributed databases. In distributed environment, the extracted rules from data mining techniques can be used in dynamic replications, adaptive load balancing and other schemes. However, transmission of large data through the system can cause errors and unreliable results. This paper proposes the intelligent distributed platform based on dynamic group binding using mobile agents which addresses the use of intelligence in distributed environment. The proposed grouping service implements classification scheme of objects. Data compressor agent and data miner agent extracts rules and compresses data, respectively, from the service node databases. The proposed algorithm performs preprocessing where it merges the less frequent dataset using neuro-fuzzy classifier before sending the data. Object group classification, data mining the service node database, data compression method, and rule extraction were simulated. Result of experiments in efficient data compression and reliable rule extraction shows that the proposed algorithm has better performance compared to other methods.

• The KIPS Transactions:PartD
• /
• v.8D no.3
• /
• pp.203-210
• /
• 2001

These days data is being collected and accumulated in a wide variety of fields. Stored data itself is to be an information system which helps us to make decisions. An information system includes many kinds of necessary and unnecessary attribute. So many algorithms have been developed for finding useful patterns from the data and reasoning approximately new objects. We are interested in the simple and understandable rules that can represent useful patterns. In this paper we propose an algorithm which can reduce the information in the system to a minimum, based on a probabilistic rough set theory. The proposed algorithm uses a value that tolerates accuracy of classification. The tolerant value helps minimizing the necessary attribute which is needed to reason a new object by reducing conditional attributes. It has the advantage that it reduces the time of generalizing rules. We experiment a proposed algorithm with the IRIS data and Wisconsin Breast Cancer data. The experiment results show that this algorithm retrieves a small reduct, and minimizes the size of the rule under the tolerant classification rate.

• Journal of KIISE:Information Networking
• /
• v.32 no.4
• /
• pp.543-550
• /
• 2005

Demand for better services in the Internet has been increasing due to the rapid growth of the Internet, and hence next generation routers are required to perform intelligent packet classification. For a given classifier defining packet attributes or contents, packet classification is the process of identifying the highest priority rule to which a packet conforms. A notable characteristic of real classifiers is that a packet matches only a small number of distinct source-destination prefix pairs. Therefore, a lot of schemes have been proposed to filter rules based on source and destination prefix pairs. However, most of the schemes are based on sequential one-dimensional searches using trio which requires huge memory. In this paper, we proposea memory-efficient two-dimensional search scheme using source and destination prefix pairs. By constructing binary prefix tree, source prefix search and destination prefix search are simultaneously performed in a binary tree. Moreover, the proposed two-dimensional binary prefix tree does not include any empty internal nodes, and hence memory waste of previous trio-based structures is completely eliminated.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.6 no.1
• /
• pp.12-23
• /
• 2003

The expansion of internet and the development of communication technology have brought about an explosive increasement of data. Further progress has led to the increasing demand for efficient and effective data analysis tools. According to this demand, data mining techniques have been developed to find out knowledge from a huge amounts of raw data. This paper suggests a generalization based classification method which explores rules from real world data appearing repeatedly. Also, it analyzed the relation between weather data and forest fire, and efficiently predicted through it as a prediction model by applying the suggested generalization based classification method to forest fire data. Additionally, the proposed method can be utilized variously in the important field of real life like the analysis and prediction on natural disaster occurring repeatedly, the prediction of energy demand and so forth.

• The Korean Journal of Applied Statistics
• /
• v.17 no.3
• /
• pp.459-473
• /
• 2004

It has well known that an exhaustive search algorithm suggested by Breiman et. a1.(1984) has a trend to select the variable having relatively many possible splits as an splitting rule. We propose an algorithm to overcome this variable selection bias problem and then construct unbiased regression trees based on the algorithm. The proposed algorithm runs two steps of selecting a split variable and determining a split rule for binary split based on the split variable. Simulation studies were performed to compare the proposed algorithm with Breiman et a1.(1984)'s CART(Classification and Regression Tree) in terms of degree of variable selection bias, variable selection power, and MSE(Mean Squared Error). Also, we illustrate the proposed algorithm with real data sets.

• Journal of Information Technology Services
• /
• v.20 no.3
• /
• pp.75-86
• /
• 2021

As artificial intelligence is actively studied, chatbot systems are being applied to various fields. In particular, many chatbot systems for psychological counseling have been studied that can comfort modern people. However, while most psychological counseling chatbots are studied as rule-base and deep learning-based chatbots, there are large limitations for each chatbot. To overcome the limitations of psychological counseling using such chatbots, we proposes a novel psychological counseling AI chatbot system. The proposed system consists of a GPT-2 model that generates output sentence for Korean input sentences and an Electra model that serves as sentiment analysis and anxiety cause classification, which can be provided with psychological tests and collective intelligence functions. At the same time as deep learning-based chatbots and conversations take place, sentiment analysis of input sentences simultaneously recognizes user's emotions and presents psychological tests and collective intelligence solutions to solve the limitations of psychological counseling that can only be done with chatbots. Since the role of sentiment analysis and anxiety cause classification, which are the links of each function, is important for the progression of the proposed system, we experiment the performance of those parts. We verify the novelty and accuracy of the proposed system. It also shows that the AI chatbot system can perform counseling excellently.