• Journal of KIISE:Software and Applications
• /
• v.29 no.4
• /
• pp.270-277
• /
• 2002

Text filtering is a task of deciding whether a document has relevance to a specified topic. As Internet and Web becomes wide-spread and the number of documents delivered by e-mail explosively grows the importance of text filtering increases as well. The aim of this paper is to improve the accuracy of text filtering systems by using machine learning techniques. We apply AdaBoost algorithms to the filtering task. An AdaBoost algorithm generates and combines a series of simple hypotheses. Each of the hypotheses decides the relevance of a document to a topic on the basis of whether or not the document includes a certain word. We begin with an existing AdaBoost algorithm which uses weak hypotheses with their output of 1 or -1. Then we extend the algorithm to use weak hypotheses with real-valued outputs which was proposed recently to improve error reduction rates and final filtering performance. Next, we attempt to achieve further improvement in the AdaBoost's performance by first setting weights randomly according to the continuous Poisson distribution, executing AdaBoost, repeating these steps several times, and then combining all the hypotheses learned. This has the effect of mitigating the ovefitting problem which may occur when learning from a small number of data. Experiments have been performed on the real document collections used in TREC-8, a well-established text retrieval contest. This dataset includes Financial Times articles from 1992 to 1994. The experimental results show that AdaBoost with real-valued hypotheses outperforms AdaBoost with binary-valued hypotheses, and that AdaBoost iterated with random weights further improves filtering accuracy. Comparison results of all the participants of the TREC-8 filtering task are also provided.

• Journal of the Korea Convergence Society
• /
• v.12 no.8
• /
• pp.279-286
• /
• 2021

The purpose of this study is to compare the predictive power of the Bagging and Boosting algorithm of ensemble method based on the motion information that occurs in woman handball matches and to analyze the availability of motion information. To this end, this study analyzed the predictive power of the result of 15 practice matches based on inertial motion by analyzing the predictive power of Random Forest and Adaboost algorithms. The results of the study are as follows. First, the prediction rate of the Random Forest algorithm was 66.9 ± 0.1%, and the prediction rate of the Adaboost algorithm was 65.6 ± 1.6%. Second, Random Forest predicted all of the winning results, but none of the losing results. On the other hand, the Adaboost algorithm shows 91.4% prediction of winning and 10.4% prediction of losing. Third, in the verification of the suitability of the algorithm, the Random Forest had no overfitting error, but Adaboost showed an overfitting error. Based on the results of this study, the availability of motion information is high when predicting sports events, and it was confirmed that the Random Forest algorithm was superior to the Adaboost algorithm.

• The KIPS Transactions:PartB
• /
• v.12B no.3 s.99
• /
• pp.349-356
• /
• 2005

Artificial Organism-used application areas are expanding at a break-neck speed with a view to getting things done in a dynamic and Informal environment. A use of general programming or traditional hi methods as the representation of Artificial Organism behavior knowledge in these areas can cause problems related to frequent modifications and bad response in an unpredictable situation. Strategies aimed at solving these problems in a machine-learning fashion includes Genetic Programming and Evolving Neural Networks. But the learning method of Artificial-Organism is not good yet, and can't represent life in the environment. With this in mind, this research is designed to come up with a new behavior evolution model. The model represents behavior knowledge with Classification Rules and Enhanced Backpropation Neural Networks and discriminate the denomination. To evaluate the model, the researcher applied it to problems with the competition of Artificial-Organism in the Simulator and compared with other system. The survey shows that the model prevails in terms of the speed and Qualify of learning. The model is characterized by the simultaneous learning of classification rules and neural networks represented on chromosomes with the help of Genetic Algorithm and the consolidation of learning ability caused by the hybrid processing of the classification rules and Enhanced Backpropagation Neural Network.

• Journal of Convergence for Information Technology
• /
• v.11 no.11
• /
• pp.57-65
• /
• 2021

We propose a novel method to detect abnormal data of specific symptoms using deep learning in air pollution measurement system. Existing methods generally detect abnomal data by classifying data showing unusual patterns different from the existing time series data. However, these approaches have limitations in detecting specific symptoms. In this paper, we use DeepLab V3+ model mainly used for foreground segmentation of images, whose structure has been changed to handle one-dimensional data. Instead of images, the model receives time-series data from multiple sensors and can detect data showing specific symptoms. In addition, we improve model's performance by reducing the complexity of noisy form time series data by using 'piecewise aggregation approximation'. Through the experimental results, it can be confirmed that anomaly data detection can be performed successfully.

• The Korean Journal of Applied Statistics
• /
• v.34 no.3
• /
• pp.309-327
• /
• 2021

Dynamic treatment regimes (DTRs) are decision-making rules designed to provide personalized treatment to individuals in multi-stage randomized trials. Unlike classical methods, in which all individuals are prescribed the same type of treatment, DTRs prescribe patient-tailored treatments which take into account individual characteristics that may change over time. The Q-learning method, one of regression-based algorithms to figure out optimal treatment rules, becomes more popular as it can be easily implemented. However, the performance of the Q-learning algorithm heavily relies on the correct specification of the Q-function for response, especially in observational studies. In this article, we examine a number of double-robust weighted least-squares estimating methods for Q-learning in high-dimensional settings, where treatment models for propensity score and penalization for sparse estimation are also investigated. We further consider flexible ensemble machine learning methods for the treatment model to achieve double-robustness, so that optimal decision rule can be correctly estimated as long as at least one of the outcome model or treatment model is correct. Extensive simulation studies show that the proposed methods work well with practical sample sizes. The practical utility of the proposed methods is proven with real data example.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.4
• /
• pp.1887-1898
• /
• 2018

In this paper, a method to classify insider threat activity is introduced. The internal threats help detecting anomalous activity in the procedure performed by the user in an organization. When an anomalous value deviating from the overall behavior is displayed, we consider it as an inside threat for classification as an inside intimidator. To solve the situation, Markov Chain Model is employed. The Markov Chain Model shows the next state value through an arbitrary variable affected by the previous event. Similarly, the current activity can also be predicted based on the previous activity for the insider threat activity. A method was studied where the change items for such state are defined by a transition probability, and classified as detection of anomaly of the inside threat through values for a probability variable. We use the properties of the Markov chains to list the behavior of the user over time and to classify which state they belong to. Sequential data sets were generated according to the influence of n occurrences of Markov attribute and classified by machine learning algorithm. In the experiment, only 15% of the Cert: insider threat dataset was applied, and the result was 97% accuracy except for NaiveBayes. As a result of our research, it was confirmed that the Markov Chain Model can classify insider threats and can be fully utilized for user behavior classification.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.10
• /
• pp.5132-5148
• /
• 2017

Cyber attacks are evolving commensurate with recent developments in information security technology. Intrusion detection systems collect various types of data from computers and networks to detect security threats and analyze the attack information. The large amount of data examined make the large number of computations and low detection rates problematic. Feature selection is expected to improve the classification performance and provide faster and more cost-effective results. Despite the various feature selection studies conducted for intrusion detection systems, it is difficult to automate feature selection because it is based on the knowledge of security experts. This paper proposes a feature selection technique to overcome the performance problems of intrusion detection systems. Focusing on feature selection, the first phase of the proposed system aims at constructing a feature subset using a sequential forward floating search (SFFS) to downsize the dimension of the variables. The second phase constructs a classification model with the selected feature subset using a random forest classifier (RFC) and evaluates the classification accuracy. Experiments were conducted with the NSL-KDD dataset using SFFS-RF, and the results indicated that feature selection techniques are a necessary preprocessing step to improve the overall system performance in systems that handle large datasets. They also verified that SFFS-RF could be used for data classification. In conclusion, SFFS-RF could be the key to improving the classification model performance in machine learning.

• Journal of Welding and Joining
• /
• v.14 no.1
• /
• pp.99-106
• /
• 1996

Development of an automatic TIG welding system incorporating a vision sensor and torch control mechanism leads to an improved welding quality and greater production efficiency. The automatic welding system should be greatly restricted in its size and weight for the LNG(Liquefied Natural Gas) storage tank and also provide a unique torch rotating mechanism which keeps the torch tip in the constant position while the angle is changed continuously to maintain the welding torch substantially perpendicular to the weld line. The developed system is driven by two translation axes X, Z and one rotational axis. A moving line window method is adopted to the image recognition of the corrugated membranes with specular reflection. This method decides original laser stripe patterns in image which is affected by multi-reflection. A self-teaching algorithm, which guides the automatic welding machine with the information provided by the CCD camera without any previous learning of a reference trajectory, was developed for tracking the corrugated membrane of the LNG tank along the weld line.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.7
• /
• pp.1773-1780
• /
• 1996

KBANN (knowledge-based artificial neural network) combining the symbolic approach and the numerical approach has been shown to be more effective than other machine learning models. However KBANN doesn't have the theory refinement ability because the topology of network can't be altered dynamically. Although TopGen was proposed to extend the ability of KABNN in this respect, it also had some defects due to the link-ing of hidden nodes to input nodes and the use of beam search. The algorithm which could solve this TopGen's defects, by adding the hidden nodes linked to next layer nodes and using hill-climbing search with backtracking, is designed.

• Journal of KIISE:Software and Applications
• /
• v.35 no.8
• /
• pp.463-478
• /
• 2008

Due to the high dimensional problem, typically machine learning algorithms have relied on feature selection techniques in order to perform effective classification in microarray gene expression datasets. However, the large number of features compared to the number of samples makes the task of feature selection computationally inprohibitive and prone to errors. One of traditional feature selection approach was feature filtering; measuring one gene per one step. Then feature filtering was an univariate approach that cannot validate multivariate correlations. In this paper, we proposed a function for measuring both class separability and correlations. With this approach, we solved the problem related to feature filtering approach.