• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.7
• /
• pp.1773-1793
• /
• 2023

Automated machine learning, often referred to as "AutoML," is the process of automating the time-consuming and iterative procedures that are associated with the building of machine learning models. There have been significant contributions in this area across a number of different stages of accomplishing a data-mining task, including model selection, hyper-parameter optimization, and preprocessing method selection. Among them, preprocessing method selection is a relatively new and fast growing research area. The current work is focused on the recommendation of preprocessing methods, i.e., feature subset selection (FSS) algorithms. One limitation in the existing studies regarding FSS algorithm recommendation is the use of a single learner for meta-modeling, which restricts its capabilities in the metamodeling. Moreover, the meta-modeling in the existing studies is typically based on a single group of data characterization measures (DCMs). Nonetheless, there are a number of complementary DCM groups, and their combination will allow them to leverage their diversity, resulting in improved meta-modeling. This study aims to address these limitations by proposing an architecture for preprocess method selection that uses ensemble learning for meta-modeling, namely AutoFE-Sel. To evaluate the proposed method, we performed an extensive experimental evaluation involving 8 FSS algorithms, 3 groups of DCMs, and 125 datasets. Results show that the proposed method achieves better performance compared to three baseline methods. The proposed architecture can also be easily extended to other preprocessing method selections, e.g., noise-filter selection and imbalance handling method selection.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.3
• /
• pp.763-768
• /
• 2023

In this paper, the existing statistical method (ARIMA) and machine learning method (Informer) were developed and compared to predict the distribution volume of pharmaceuticals. It was found that a machine learning-based model is advantageous for daily data prediction, and it is effective to use ARIMA for monthly prediction and switch to Informer as the data increases. The prediction error rate (RMSE) was reduced by 26.6% compared to the previous method, and the prediction accuracy was improved by 13%, resulting in a result of 86.2%. Through this thesis, we find that there is an advantage of obtaining the best results by ensembleing statistical methods and machine learning methods. In addition, machine learning-based AI models can derive the best results through deep learning operations even in irregular situations, and after commercialization, performance is expected to improve as the amount of data increases.

• 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.

• Journal of Intelligence and Information Systems
• /
• v.15 no.3
• /
• pp.1-15
• /
• 2009

In a classification problem, data imbalance occurs when the number of instances in one class greatly outnumbers the number of instances in the other class. Such data sets often cause a default classifier to be built due to skewed boundary and thus the reduction in the classification accuracy of such a classifier. This paper proposes a Geometric Mean-based Boosting (GM-Boost) to resolve the problem of data imbalance. Since GM-Boost introduces the notion of geometric mean, it can perform learning process considering both majority and minority sides, and reinforce the learning on misclassified data. An empirical study with bankruptcy prediction on Korea companies shows that GM-Boost has the higher classification accuracy than previous methods including Under-sampling, Over-Sampling, and AdaBoost, used in imbalanced data and robust learning performance regardless of the degree of data imbalance.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.4
• /
• pp.169-178
• /
• 2017

Recently, as the interest rate and necessity for security is growing, the demands for a visitor recognition system are being increased. In order to recognize a visitor in visitor recognition systems, the various biometric methods are used. In this paper, we propose a visitor recognition system based on face recognition. The visitor recognition system improves the face recognition performance by integrating several open APIs as a single algorithm and by performing the ensemble of the recognition results. For the performance evaluation, we collected the face data for about five months and measured the performance of the visitor recognition system. As the results of the performance measurement, the visitor recognition system shows a higher face recognition rate than using a single face recognition API, meeting the requirements on performance.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1393-1399
• /
• 2019

With the rapid growth of the Internet, millions of indiscriminate advertising SMS are sent every day because of the convenience of sending and receiving data. Although we still use methods to block spam words manually, we have been actively researching how to filter spam in a various ways as machine learning emerged. However, spam words and patterns are constantly changing to avoid being filtered, so existing machine learning mechanisms cannot detect or adapt to new words and patterns. Recently, the concept of Lifelong Learning emerged to overcome these limitations, using existing knowledge to keep learning new knowledge continuously. In this paper, we propose a method of spam filtering system using ensemble techniques of naive bayesian which is most commonly used in document classification and LLML(Lifelong Machine Learning). We validate the performance of lifelong learning by applying the model ELLA and the Naive Bayes most commonly used in existing spam filters.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.4
• /
• pp.13-29
• /
• 2020

Automobiles have long been an essential part of daily life, but the social costs of car traffic accidents exceed 9% of the national budget of Korea. Hence, it is necessary to establish prevention and response system for car traffic accidents. In order to present a model that can classify and predict the degree of injury in car traffic accidents, we used big data analysis techniques of K-nearest neighbor, logistic regression analysis, naive bayes classifier, decision tree, and ensemble algorithm. The performances of the models were analyzed by using the data on the nationwide traffic accidents over the past three years. In particular, considering the difference in the number of data among the respective injury severity levels, we used down-sampling methods for the group with a large number of samples to enhance the accuracy of the classification of the models and then verified the statistical significance of the models using ANOVA.

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.3
• /
• pp.45-54
• /
• 2014

In this paper, an improved two-step P2P traffic classification scheme is proposed to overcome the limitations of the existing methods. The first step is a signature-based classifier at the packet-level. The second step consists of pattern heuristic rules and a statistics-based classifier at the flow-level. With pattern heuristic rules, the accuracy can be improved and the amount of traffic to be classified by statistics-based classifier can be reduced. Based on the analysis of different decision tree algorithms, the statistics-based classifier is implemented with REPTree. In addition, the ensemble algorithm is used to improve the performance of statistics-based classifier Through the verification with the real datasets, it is shown that our hybrid scheme provides higher accuracy and lower overhead compared to other existing schemes.

• Journal of Korean Society for Quality Management
• /
• v.48 no.3
• /
• pp.493-510
• /
• 2020

Purpose: As of July 31, there were 14,336 confirmed cases of COVID-19 in South Korea, including 301 deaths. Since the daily confirmed number of cases hit 909 on February 29, the spread of the disease had gradually decreased due to the active implementation of preventive control interventions, and the daily confirmed number had finally recorded a single digit on April 19. Since May, however, the disease has re-emerged and retaining after June. In order to eradicate the disease, it is necessary to suggest suitable forward preventive strategies by predicting future infectivity of the disease based on the cases so far. Therefore, in this study, we aim to evaluate the transmission potential of the disease in early phases by estimating basic reproduction number and assess the preventive control measures through effective reproduction number. Methods: We used publicly available cases and deaths data regarding COVID-19 in South Korea as of July 31. Using ensemble model integrated stochastic linear birth model and deterministic linear growth model, the basic reproduction number and the effective reproduction number were estimated. Results: Estimated basic reproduction number is 3.1 (95% CI: 3.0-3.2). Effective reproduction number was the highest with 7 on February 15, decreased as of April 20. Since then, the value is gradually increased to more than unity. Conclusion: Preventive policy such as wearing a mask and physical distancing campaigns in the early phase of the outbreak was fairly implemented. However, the infection potential increased due to weakening government policy on May 6. Our results suggest that it seems necessary to implement a stronger policy than the current level.

• Korean Journal of Health Education and Promotion
• /
• v.23 no.4
• /
• pp.13-28
• /
• 2006

Objectives: This study used the characteristics of the knowledge discovery and data mining algorithms to develop hypertension predictive model for hypertension management using the Korea National Health Insurance Corporation database(the insureds' screening and health care benefit data). Methods: This study validated the predictive power of data mining algorithms by comparing the performance of logistic regression, decision tree, and ensemble technique. On the basis of internal and external validation, it was found that the model performance of logistic regression method was the best among the above three techniques. Results: Major results of logistic regression analysis suggested that the probability of hypertension was: - lower for the female(compared with the male)(OR=0.834) - higher for the persons whose ages were 60 or above(compared with below 40)(OR=4.628) - higher for obese persons(compared with normal persons)(OR= 2.103) - higher for the persons with high level of glucose(compared with normal persons)(OR=1.086) - higher for the persons who had family history of hypertension(compared with the persons who had not)(OR=1.512) - higher for the persons who periodically drank alcohol(compared with the persons who did not)$(OR=1.037{\sim}1.291)$ Conclusions: This study produced several factors affecting the outbreak of hypertension using screening. It is considered to be a contributing factor towards the nation's building of a Hypertension Management System in the near future by bringing forth representative results on the rise and care of hypertension.