• 디지털산업정보학회논문지
• /
• 제6권1호
• /
• pp.1-12
• /
• 2010

Recently, large amount of information in IDS(Intrusion Detection System) can be un manageable and also be mixed with false prediction error. In this paper, we propose a data mining methodology for IDS, which contains uncertainty based on training process and post-processing analysis additionally. Our system is trained to classify the existing attack for misuse detection, to detect the new attack pattern for anomaly detection, and to define border patter between attack and normal pattern. In experimental results show that our approach improve the performance against existing attacks and new attacks,from 0.62 to 0.84 about 35%.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2022년도 춘계학술대회
• /
• pp.277-279
• /
• 2022

Rapid Development and adoption of AIS as a survailance tool has resulted in widespread application of data analysis technology, in addition to AIS ship trajectory clustering. AIS data-based clustering has become an increasingly popular method for marine traffic pattern recognition, ship route prediction and anomaly detection in recent year. In this paper we propose a route waypoint extraction by clustering ships CoG variance trajectory using Density-Based Spatial Clustering of Application with Noise (DBSCAN) algorithm in both port approach channel and coastal waters. The algorithm discovers route waypoint effectively. The result of the study could be used in traffic route extraction, and more-so develop a maritime anomaly detection tool.

• 대기
• /
• 제24권4호
• /
• pp.533-540
• /
• 2014

Statistical forecast models for the prediction of the summertime Changma precipitation have been developed in this study. As effective predictors for the Changma precipitation, the springtime sea surface temperature (SST) anomalies over the North Atlantic (NA1), the North Pacific (NPC) and the tropical Pacific Ocean (CNINO) has been suggested in Lee and Seo (2013). To further improve the performance of the statistical prediction scheme, we select other potential predictors and construct 2 additional statistical models. The selected predictors are the Northern Indian Ocean (NIO) and the Bering Sea (BS) SST anomalies, and the spring Eurasian snow cover anomaly (EUSC). Then, using the total three statistical prediction models, a simple ensemble-mean prediction is performed. The resulting correlation skill score reaches as high as ~0.90 for the last 21 years, which is ~16% increase in the skill compared to the prediction model by Lee and Seo (2013). The EUSC and BS predictors are related to a strengthening of the Okhotsk high, leading to an enhancement of the Changma front. The NIO predictor induces the cyclonic anomalies to the southwest of the Korean peninsula and southeasterly flows toward the peninsula, giving rise to an increase in the Changma precipitation.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권5호
• /
• pp.1690-1707
• /
• 2021

Precipitation prediction during flood season has been a key task of climate prediction for a long time. This type of prediction is linked with the national economy and people's livelihood, and is also one of the difficult problems in climatology. At present, there are some precipitation forecast models for the flood season, but there are also some deviations from these models, which makes it difficult to forecast accurately. In this paper, based on the measured precipitation data from the flood season from 1993 to 2019 and the precipitation return data of CWRF, ANN cycle modeling and a weighted integration method is used to correct the CWRF used in today's operational systems. The MAE and TCC of the precipitation forecast in the flood season are used to check the prediction performance of the proposed algorithm model. The results demonstrate a good correction effect for the proposed algorithm. In particular, the MAE error of the new algorithm is reduced by about 50%, while the time correlation TCC is improved by about 40%. Therefore, both the generalization of the correction results and the prediction performance are improved.

• 대기
• /
• 제26권1호
• /
• pp.203-214
• /
• 2016

This study explores the 6-month lead prediction skill of stratospheric temperature and circulations in the Global Seasonal forecasting model version 5 (GloSea5) hindcast experiment over the period of 1996~2009. Both the tropical and extratropical circulations are considered by analyzing the Quasi-Biennial Oscillation (QBO) and Northern Hemisphere Polar Vortex (NHPV). Their prediction skills are quantitatively evaluated by computing the Anomaly Correlation Coefficient (ACC) and Mean Squared Skill Score (MSSS), and compared with those of El Nino-Southern Oscillation (ENSO) and Arctic Oscillation (AO). Stratospheric temperature is generally better predicted than tropospheric temperature. Such improved prediction skill, however, rapidly disappears in a month, and a reliable prediction skill is observed only in the tropics, indicating a higher prediction skill in the tropics than in the extratropics. Consistent with this finding, QBO is well predicted more than 6 months in advance. Its prediction skill is significant in all seasons although a relatively low prediction skill appears in the spring when QBO phase transition often takes place. This seasonality is qualitatively similar to the spring barrier of ENSO prediction skill. In contrast, NHPV exhibits no prediction skill beyond one month as in AO prediction skill. In terms of MSSS, both QBO and NHPV are better predicted than their counterparts in the troposphere, i.e., ENSO and AO, indicating that the GloSea5 has a higher prediction skill in the stratosphere than in the troposphere.

• 한국터널지하공간학회 논문집
• /
• 제26권5호
• /
• pp.519-532
• /
• 2024

TBM (tunnel boring machine) 터널 프로젝트의 리스크 관리 측면에서 굴진율 예측은 중요하며, 이를 위한 머신러닝 기반 TBM 굴진율 예측 연구가 지속적으로 진행되어 왔다. 그러나, 기존 연구의 머신러닝 예측 모델은 정상 굴진율과 이상 굴진율 간의 불균형 데이터를 고려하는 데 한계가 있다. 본 연구에서는 데이터 증강 기법을 통해 불균형 데이터를 처리하여 머신러닝 기반 TBM 굴진율 이상탐지 성능을 개선하였다. 먼저, 상관관계 분석을 통해 유사 변수를 제거하여 6가지 입력특성을 선정하였다. 또한, 하위 10%와 상위 10%의 굴진율을 각각 이상 등급으로, 그 외 범위의 굴진율을 정상 등급으로 굴진율 등급을 구분하였다. 기존 학습 데이터와 SMOTE (synthetic minority oversampling technique)를 통해 증강된 학습 데이터를 각각 XGB (extreme gradient boosting)에 적용한 XGB 모델과 XGB-SMOTE 모델을 구축하였다. 굴진율 등급 예측 성능을 비교한 결과, XGB 모델은 정상 굴진율에 대한 예측 성능은 우수하나 이상 굴진율 예측 성능은 상대적으로 낮게 도출되었다. 반면, XGB-SMOTE 모델은 모든 굴진율 등급에서 일관되게 우수한 예측 성능을 보였다. 이는 SMOTE를 통한 이상 굴진율 데이터의 증강이 이상 굴진율을 유발하는 지반조건과 TBM 운영인자 간의 패턴 학습 수준을 향상시켰기 때문으로 판단된다. 결론적으로, 본 연구는 머신러닝 기반 TBM 굴진율 이상탐지 시 데이터 증강 기법을 활용한 불균형 데이터 처리가 효과적임을 보여준다.

• 대기
• /
• 제28권3호
• /
• pp.233-245
• /
• 2018

This study explores the Subseasonal-to-Seasonal (S2S) prediction skills of the Northern Hemisphere mid-latitude geopotential height in the Global Seasonal forecasting model version 5 (GloSea5) hindcast experiment. The prediction skills are quantitatively verified for the period of 1991~2010 by computing the Anomaly Correlation Coefficient (ACC) and Mean Square Skill Score (MSSS). GloSea5 model shows a higher prediction skill in winter than in summer at most levels regardless of verification methods. Quantitatively, the prediction limit diagnosed with ACC skill of 500 hPa geopotential height, averaged over $30^{\circ}N{\sim}90^{\circ}N$, is 11.0 days in winter, but only 9.1 days in summer. These prediction limits are primarily set by the planetary-scale eddy phase errors. The stratospheric prediction skills are typically higher than the tropospheric skills except in the summer upper-stratosphere where prediction skills are substantially lower than upper-troposphere. The lack of the summer upper-stratospheric prediction skill is caused by zonal mean error, perhaps strongly related to model mean bias in the stratosphere.

• 한국산학기술학회논문지
• /
• 제17권9호
• /
• pp.440-446
• /
• 2016

군사 비밀이나 조직의 기밀 데이터는 그 조직의 매우 중요한 자원이며 외부로부터의 접근이 차단되어야 한다. 그러나 최근 인터넷의 접근성이 높아짐으로써 보안이 중요한 이슈로 부상하고 있다. 이를 위해 네트워크 내부에 대한 공격이나 침입행위를 탐지하는 이상 행위 탐지 방법이 제안되었다. 그러나 대부분의 이상 행위 탐지는 외부로부터의 침입에 대한 측면만 다루고 있으며, 공격이나 침입보다 더 큰 피해를 입히는 내부 데이터의 유출에 대해서는 다루고 있지 않다. 또한 기존의 이상 행위 탐지 방법을 데이터 유출 탐지에 적용할 경우 네트워크 내부의 환경과 여러 가지 변수들이 고려되어 있지 않기 때문에 많은 문제점들이 발생한다. 따라서 본 논문에서는 데이터 유출 탐지를 위한 이상 행위 탐지(Data Exfiltrating Detection for Anomaly Detection : DEDfAD) 방법의 정확도 향상을 위하여 DEDfAD에서 고려되어야 하는 이슈 사항들에 대하여 기술하고, 프로파일 기반의 탐지 방법과 머신러닝 기반의 탐지 방법으로 분류하여 이들의 장단점을 분석한다. 또한 분류된 접근 방법을 중심으로 이슈들과의 비교분석을 통해 향후 연구 방향을 제시한다.

• Journal of Korea Artificial Intelligence Association
• /
• 제1권1호
• /
• pp.11-16
• /
• 2023

Accurate hospital case modeling and prediction are crucial for efficient healthcare. In this study, we demonstrate the implementation of regression analysis methods in machine learning systems utilizing mathematical statics and machine learning techniques. The developed machine learning model includes Bayesian linear, artificial neural network, decision tree, decision forest, and linear regression analysis models. Through the application of these algorithms, corresponding regression models were constructed and analyzed. The results suggest the potential of leveraging machine learning systems for medical research. The experiment aimed to create an Azure Machine Learning Studio tool for the speedy evaluation of multiple regression models. The tool faciliates the comparision of 5 types of regression models in a unified experiment and presents assessment results with performance metrics. Evaluation of regression machine learning models highlighted the advantages of boosted decision tree regression, and decision forest regression in hospital case prediction. These findings could lay the groundwork for the deliberate development of new directions in medical data processing and decision making. Furthermore, potential avenues for future research may include exploring methods such as clustering, classification, and anomaly detection in healthcare systems.

• 대기
• /
• 제34권2호
• /
• pp.83-95
• /
• 2024

Atmospheric rivers (ARs) are closely related to local precipitation which can be both beneficial and destructive. Although several studies have evaluated their predictability, there is a lack of studies on East Asian ARs. This study evaluates the prediction skill of East Asian ARs in the Korean Integrated Model (KIM) for 2020~2022 summer. The spatial distribution of AR frequency in KIM is qualitatively similar to the observation but overestimated. In particular, the model errors greatly increase along the boundary of the western North Pacific subtropical high as the forecast lead time increases. When the prediction skills are quantitatively verified by computing the Anomaly Correlation Coefficient and Mean Square Skill Score, the useful prediction skill of daily AR around the Korean Peninsula is found up to 5 days. Such prediction limit is primarily set by the wind field errors with a minor contribution of moisture distribution errors. This result suggests that the improved prediction of atmospheric circulation field can improve the prediction of East Asian summer ARs and the associated precipitation.