Sanggil Kang

Prof. Sanggil Kang


Research Focus & Metrics

Academic Career Length

19Years (2008-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
6 September 2017 in Journal of Healthcare Engineering
Verified
Journal Article
Open access
PDF file
JKJae Kwon Kim
Jae Kwon Kim
SKSanggil Kang
Sanggil Kang
Background. Of the machine learning techniques used in predicting coronary heart disease (CHD), neural network (NN) is popularly used to improve performance accuracy. Objective. Even though NN-based systems provide meaningful results based on clinical experiments, medical experts are not satisfied with their predictive performances because NN is trained in a “black-box” style. Method. We sought to devise an NN-based prediction of CHD risk using feature correlation analysis (NN-FCA) using two stages. First, the feature selection stage, which makes features acceding to the importance in predicting CHD risk, is ranked, and second, the feature correlation analysis stage, during which one learns about the existence of correlations between feature relations and the data of each NN predictor output, is determined. Result. Of the 4146 individuals in the Korean dataset evaluated, 3031 had low CHD risk and 1115 had CHD high risk. The area under the receiver operating characteristic (ROC) curve of the proposed model (0.749 ± 0.010) was larger than the Framingham risk score (FRS) (0.393 ± 0.010). Conclusions. The proposed NN-FCA, which utilizes feature correlation analysis, was found to be better than FRS in terms of CHD risk prediction. Furthermore, the proposed model resulted in a larger ROC curve and more accurate predictions of CHD risk in the Korean population than the FRS.

22 April 2022 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
HHHyeontaek Hwang
Hyeontaek Hwang
SKSanggil Kang
Sanggil Kang
Many non-intrusive load monitoring (NILM) studies use high-frequency data to classify the device’s ON/OFF state. However, these approaches cannot be applied in real-world situations due to increased network traffic and database capacity issues. For these reasons, when trying to perform NILM with low-frequency data, the power usage pattern that changes over time disappears and features cannot be properly obtained to classify devices. In this article, we propose a novel NILM model that can learn datasets with imbalanced data classes. The model extracts features using long-short term memory (LSTM) and improve the feature representation ability of LSTM through the feedback of predictions. The experiment is conducted using the REDD dataset and the Living-lab validation dataset. In the REDD dataset, the proposed method outperforms conventional methods 10%–20% on the Majority device and 50%–60% on the Minority device. Living-lab validation results show that the performance of the proposed method outperforms other previously proposed NILM systems in low-frequency data and can be applied to real-world NILM situations.

20 June 2008 in Pattern Recognition Letters
Verified
Journal Article
Closed access
SKSanggil Kang
Sanggil Kang
SPSungjoon Park
Sungjoon Park

10 November 2018 in The Journal of Supercomputing
Verified
Journal Article
Closed access
PDF file
CHCheonghwan Hur
Cheonghwan Hur
SKSanggil Kang
Sanggil Kang

1 November 2010 in Sensors
Verified
Journal Article
Open access
PDF file
Yujin LimYujin Lim
Yujin Lim
HKHak-Man Kim
Hak-Man Kim
SKSanggil Kang
Sanggil Kang
Power grids deal with the business of generation, transmission, and distribution of electric power. Recently, interest in power quality in electrical distribution systems has increased rapidly. In Korea, the communication network to deliver voltage, current, and temperature measurements gathered from pole transformers to remote monitoring centers employs cellular mobile technology. Due to high cost of the cellular mobile technology, power quality monitoring measurements are limited and data gathering intervals are large. This causes difficulties in providing the power quality monitoring service. To alleviate the problems, in this paper we present a communication infrastructure to provide low cost, reliable data delivery. The communication infrastructure consists of wired connections between substations and monitoring centers, and wireless connections between pole transformers and substations. For the wireless connection, we employ a wireless sensor network and design its corresponding data forwarding protocol to improve the quality of data delivery. For the design, we adopt a tree-based data forwarding protocol in order to customize the distribution pattern of the power quality information. We verify the performance of the proposed data forwarding protocol quantitatively using the NS-2 network simulator.

4 August 2022 in Sensors
Verified
Journal Article
Open access
PDF file
CHCheong-Hwan Hur
Cheong-Hwan Hur
HLHan-Eum Lee
Han-Eum Lee
YKYoung-Joo Kim
Young-Joo Kim
SKSanggil Kang
Sanggil Kang
Nonintrusive load monitoring (NILM) is a technology that analyzes the load consumption and usage of an appliance from the total load. NILM is becoming increasingly important because residential and commercial power consumption account for about 60% of global energy consumption. Deep neural network-based NILM studies have increased rapidly as hardware computation costs have decreased. A significant amount of labeled data is required to train deep neural networks. However, installing smart meters on each appliance of all households for data collection requires the cost of geometric series. Therefore, it is urgent to detect whether the appliance is used from the total load without installing a separate smart meter. In other words, domain adaptation research, which can interpret the huge complexity of data and generalize information from various environments, has become a major challenge for NILM. In this research, we optimize domain adaptation by employing techniques such as robust knowledge distillation based on teacher–student structure, reduced complexity of feature distribution based on gkMMD, TCN-based feature extraction, and pseudo-labeling-based domain stabilization. In the experiments, we down-sample the UK-DALE and REDD datasets as in the real environment, and then verify the proposed model in various cases and discuss the results.