• Korean Journal of Radiology
• /
• 제20권8호
• /
• pp.1275-1284
• /
• 2019

Objective: To develop algorithms using convolutional neural networks (CNNs) for automatic segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) and compare them with conventional algorithms, including a thresholding-based segmentation. Materials and Methods: Between September 2005 and August 2015, 429 patients presenting with acute cerebral ischemia (training:validation:test set = 246:89:94) were retrospectively enrolled in this study, which was performed under Institutional Review Board approval. Ground truth segmentations for acute ischemic lesions on DWI were manually drawn under the consensus of two expert radiologists. CNN algorithms were developed using two-dimensional U-Net with squeeze-and-excitation blocks (U-Net) and a DenseNet with squeeze-and-excitation blocks (DenseNet) with squeeze-and-excitation operations for automatic segmentation of acute ischemic lesions on DWI. The CNN algorithms were compared with conventional algorithms based on DWI and the apparent diffusion coefficient (ADC) signal intensity. The performances of the algorithms were assessed using the Dice index with 5-fold cross-validation. The Dice indices were analyzed according to infarct volumes (< 10 mL, ≥ 10 mL), number of infarcts (≤ 5, 6-10, ≥ 11), and b-value of 1000 (b1000) signal intensities (< 50, 50-100, > 100), time intervals to DWI, and DWI protocols. Results: The CNN algorithms were significantly superior to conventional algorithms (p < 0.001). Dice indices for the CNN algorithms were 0.85 for U-Net and DenseNet and 0.86 for an ensemble of U-Net and DenseNet, while the indices were 0.58 for ADC-b1000 and b1000-ADC and 0.52 for the commercial ADC algorithm. The Dice indices for small and large lesions, respectively, were 0.81 and 0.88 with U-Net, 0.80 and 0.88 with DenseNet, and 0.82 and 0.89 with the ensemble of U-Net and DenseNet. The CNN algorithms showed significant differences in Dice indices according to infarct volumes (p < 0.001). Conclusion: The CNN algorithm for automatic segmentation of acute ischemic lesions on DWI achieved Dice indices greater than or equal to 0.85 and showed superior performance to conventional algorithms.

• 한국융합학회논문지
• /
• 제12권8호
• /
• pp.279-286
• /
• 2021

본 연구는 여자핸드볼 경기에서 발생되는 움직임 정보를 바탕으로 앙상블 기법의 배깅과 부스팅 알고리즘의 예측력을 비교하고, 움직임 정보의 활용가능성을 분석하는데 목적이 있다. 연구의 목적을 달성하기 위하여 15번의 연습경기에서 관성센서를 활용해 수집한 움직임 정보를 활용한 경기 결과예측을 랜덤포레스트와 Adaboost 알고리즘을 활용해 비교·분석하였다. 연구결과 첫째, 랜덤포레스트 알고리즘의 예측률은 66.9 ± 0.1%로 나타났으며, Adaboost 알고리즘의 예측률은 65.6 ± 1.6%로 나타났다. 둘째, 랜덤포레스트는 승리 결과는 모두 예측하였고, 패배의 결과는 하나도 예측하지 못하였다. 반면, Adaboost 알고리즘은 승리 예측 91.4%, 패배예측 10.4%라고 나타났다. 셋째, 알고리즘의 적합성 여부에서 랜덤포레스트는 과적합의 오류가 없었지만, Adaboost는 과적합의 오류가 나타났다. 본 연구결과를 바탕으로 스포츠경기를 예측할 때 움직임 정보도 활용 가능성을 확인하였으며, 랜덤포레스트 알고리즘이 보다 우수함을 확인하였다.

• 대한원격탐사학회지
• /
• 제34권1호
• /
• pp.141-150
• /
• 2018

Crop type classification is essential for supporting agricultural decisions and resource monitoring. Remote sensing techniques, especially using hyperspectral imagery, have been effective in agricultural applications. Hyperspectral imagery acquires contiguous and narrow spectral bands in a wide range. However, large dimensionality results in unreliable estimates of classifiers and high computational burdens. Therefore, reducing the dimensionality of hyperspectral imagery is necessary. In this study, the Random Forest (RF) classifier was utilized for dimensionality reduction as well as classification purpose. RF is an ensemble-learning algorithm created based on the Classification and Regression Tree (CART), which has gained attention due to its high classification accuracy and fast processing speed. The RF performance for crop classification with airborne hyperspectral imagery was assessed. The study area was the cultivated area in Chogye-myeon, Habcheon-gun, Gyeongsangnam-do, South Korea, where the main crops are garlic, onion, and wheat. Parameter optimization was conducted to maximize the classification accuracy. Then, the dimensionality reduction was conducted based on RF variable importance. The result shows that using the selected bands presents an excellent classification accuracy without using whole datasets. Moreover, a majority of selected bands are concentrated on visible (VIS) region, especially region related to chlorophyll content. Therefore, it can be inferred that the phenological status after the mature stage influences red-edge spectral reflectance.

• Journal of Positioning, Navigation, and Timing
• /
• 제11권4호
• /
• pp.361-368
• /
• 2022

An indoor navigation system that utilizes long-term evolution (LTE) signals has the benefit of no additional infrastructure installation expenses and low base station database management costs. Among the LTE signal measurements, received signal strength (RSS) is particularly appealing because it can be easily obtained with mobile devices. Propagation channel models can be used to estimate the position of mobile devices with RSS. However, conventional channel models have a shortcoming in that they do not discriminate between line-of-sight (LOS) and non-line-of-sight (NLOS) conditions of the received signal. Accordingly, a previous study has suggested separated LOS and NLOS channel models. However, a method for determining LOS and NLOS conditions was not devised. In this study, a machine learning-based LOS/NLOS classification method using RSS measurements is developed. We suggest several machine-learning features and evaluate various machine-learning algorithms. As an indoor experimental result, up to 87.5% classification accuracy was achieved with an ensemble algorithm. Furthermore, the range estimation accuracy with an average error of 13.54 m was demonstrated, which is a 25.3% improvement over the conventional channel model.

• 한국재료학회지
• /
• 제33권4호
• /
• pp.164-174
• /
• 2023

The bandgap characteristics of semiconductor materials are an important factor when utilizing semiconductor materials for various applications. In this study, based on data provided by AFLOW (Automatic-FLOW for Materials Discovery), the bandgap of a semiconductor material was predicted using only the material's compositional features. The compositional features were generated using the python module of 'Pymatgen' and 'Matminer'. Pearson's correlation coefficients (PCC) between the compositional features were calculated and those with a correlation coefficient value larger than 0.95 were removed in order to avoid overfitting. The bandgap prediction performance was compared using the metrics of R² score and root-mean-squared error. By predicting the bandgap with randomforest and xgboost as representatives of the ensemble algorithm, it was found that xgboost gave better results after cross-validation and hyper-parameter tuning. To investigate the effect of compositional feature selection on the bandgap prediction of the machine learning model, the prediction performance was studied according to the number of features based on feature importance methods. It was found that there were no significant changes in prediction performance beyond the appropriate feature. Furthermore, artificial neural networks were employed to compare the prediction performance by adjusting the number of features guided by the PCC values, resulting in the best R² score of 0.811. By comparing and analyzing the bandgap distribution and prediction performance according to the material group containing specific elements (F, N, Yb, Eu, Zn, B, Si, Ge, Fe Al), various information for material design was obtained.

• 한국해양학회지:바다
• /
• 제26권4호
• /
• pp.307-326
• /
• 2021

상시 관측되는 조위관측소 해수위 자료는 결측값과 오측값을 포함하고 있으며, 그 중 오측 값은 이상값으로 분류되는 전처리 대상이다. 이러한 오측을 제거하기 위해 대표적으로 3𝜎 (three standard deviations) 규칙이 적용되어왔으나, 기상이변 등에 의한 극값이 존재하거나 3𝜎 범위 안에서도 오측이 존재하는 해수위 자료에는 그 적용이 어렵다. 본 연구에서 설계된 모델은 오측에 대한 사전 정보가 필요하지 않은 비주석 학습으로 구성되며, 재귀신경망과 앙상블 기법을 이용함으로써 실시간으로 수집되는 해수위 자료가 오측일 가능성을 발생한지 20분 이내로 제시한다. 검증이 완료된 모델은 평시 및 기상이변시의 정상값과 오측값을 잘 분리하며, 학습이 이뤄지지 않은 연도의 해수위 자료에서도 이상값 탐지가 가능함을 확인하였다. 본 연구의 관측 이상치 탐지 알고리즘은 조위관측소 해수위에 국한되지 않고 다양한 해양 및 대기자료의 이상치 탐지 인공신경망 모델에 확장 적용할 수 있다.

• 대한원격탐사학회지
• /
• 제20권4호
• /
• pp.235-244
• /
• 2004

본 연구는 SPOT-4 위성의 VEGETATION-1 센서의 가시 채널, 근적외 채널, 단파 적외채널 자료를 이용하여 눈과 구름을 구별하기 위해 새롭게 제시된 알고리즘을 평가하기 위한 것이다. 눈과 구름의 마스크를 위해 전통적으로 이용되고 있는 임계치 방법들은 본 연구에서 좋은 결과를 보여 주지 못하였다 따라서 K-means 군집화 방법이 이러한 임계치 방법 대신 본 연구에서 사용되었다. 군집화에서는 두 임계치 알고리즘을 통합하여 적설과 구름을 그룹화 시켜 동시에 추출한 화소들을 적용하였다. 이것은 전체 영상을 군집화에 적용시킬 때와 비교해 군집화의 과정을 단순화시키고 나아가 정확도를 향상시킬 수 있다. 본 연구는 이러한 과정을 통해 얻어진 결과를 임계치 방법이 적용되었을 때의 결과와 비교함과 동시에 VEGETATION 자료의 분별능력을 평가하였다. 본 연구에서 제시한 방법을 이용하였을 때, 구름과 눈의 분별 능력은 상당히 향상되었다. 분별 오차는 임계치 방법을 사용하였을 때 보다 구름에 대해 19.4% 적설에 대해 9.7% 정도 감소하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권7호
• /
• 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.

• 한국컴퓨터정보학회논문지
• /
• 제19권3호
• /
• pp.45-54
• /
• 2014

본 논문에서는 기존 기법들의 제한점을 개선하기 위해 휴리스틱 규칙 및 기계학습 분석 결과를 이용한 두 단계의 P2P 트래픽 분류 기법을 제안한다. 첫 번째 단계는 패킷 레벨의 시그니처 기반 분류기이고, 두 번째 단계는 플로우 레벨에서 수행되는 패턴 휴리스틱 규칙 및 통계 기반 분류기이다. 제안된 패턴 휴리스틱 규칙은 분류의 정확도를 높이고 통계 기반 분류기가 처리할 트래픽의 양을 줄일 수 있다. 다양한 의사 결정 트리 알고리즘의 분석을 기반으로 통계 기반 분류기는 가장 효율적인 REPTree로 구현하고, 앙상블 알고리즘을 통해 통계 기반 분류기의 성능을 개선한다. 실제 환경의 데이터 집합을 이용한 검증 분석을 통해, 본 제안 기법이 기존 기법에 비해 높은 정확도와 낮은 과부하를 제공함을 제시한다.

• 방송공학회논문지
• /
• 제8권2호
• /
• pp.153-162
• /
• 2003

본 논문은 우리나라의 지상파 디지털 TV방식인 ATSC (Advanced Television System Committee) 8-VSB (Vestigial Sideband) 시스템의 수신부 가운데 등화부에 해당되는 DFE (Decision Feedback Equalizer)에 관한 것으로서, DFE 구조의 전반적인 개요와 블라인드 알고리즘에 대한 성능 분석결과를 서술한다. 특히, 다중경로, 도플러 천이(Doppler Shift), 건물 벽에 의한 신호의 감쇄 등의 영향으로 수신이 어려운 실내수신 환경에서의 수신 성능개선을 위해 최적화된 등화기의 구조를 제시하고, 등화기와 TCM(Trellis Coded Modulation)연동방법 및 필터계수 초기화 알고리즘 등을 구현한 후 모의 실험을 통한 성능 분석결과를 제시한다.