• ETRI Journal
• /
• v.46 no.5
• /
• pp.793-805
• /
• 2024

Entanglement is crucial for achieving quantum advantages. However, in the context of quantum machine learning, existing optimization strategies for generating quantum classifier circuits often result in unentangled circuits, indicating an underutilization of the entanglement effect needed to learn complex patterns. In this study, we proposed a novel metaheuristic approach-genetic algorithm-for designing a quantum kernel classifier that incorporates expressive entanglement. This classifier utilizes a loopless entanglement-directed graph, where each directed edge represents the entanglement between the target and control qubits. The proposed method consistently outperforms classical and quantum baselines across various artificial and actual datasets, achieving improvements up to 32.4% and 17.5%, respectively, compared with the best model among all other baselines. Moreover, this method successfully reconstructs the hidden entanglement structures underlying artificial datasets. The results also demonstrate that the optimized circuits exhibit diverse entanglement variations across different datasets, indicating the versatility of the proposed approach.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.5
• /
• pp.274-286
• /
• 2022

Multi-object tracking has been studied for a long time under computer vision and plays a critical role in applications such as autonomous driving and driving assistance. Multi-object tracking techniques generally consist of a detector that detects objects and a tracker that tracks the detected objects. Various publicly available datasets allow us to train a detector model without much effort. However, there are relatively few publicly available datasets for training a tracker model, and configuring own tracker datasets takes a long time compared to configuring detector datasets. Hence, the detector is often developed separately with a tracker module. However, the separated tracker should be adjusted whenever the former detector model is changed. This study proposes a system that can train a model that performs detection and tracking simultaneously using only the detector training datasets. In particular, a Siam network with augmentation is used to compose the detector and tracker. Experiments are conducted on public datasets to verify that the proposed algorithm can formulate a real-time multi-object tracker comparable to the state-of-the-art tracker models.

• Journal of Internet Computing and Services
• /
• v.24 no.4
• /
• pp.127-135
• /
• 2023

This research paper explores the application of FinBERT, a variational BERT-based model pre-trained on financial domain, for sentiment analysis in the financial domain while focusing on the process of identifying suitable training data and hyperparameters. Our goal is to offer a comprehensive guide on effectively utilizing the FinBERT model for accurate sentiment analysis by employing various datasets and fine-tuning hyperparameters. We outline the architecture and workflow of the proposed approach for fine-tuning the FinBERT model in this study, emphasizing the performance of various datasets and hyperparameters for sentiment analysis tasks. Additionally, we verify the reliability of GPT-3 as a suitable annotator by using it for sentiment labeling tasks. Our results show that the fine-tuned FinBERT model excels across a range of datasets and that the optimal combination is a learning rate of 5e-5 and a batch size of 64, which perform consistently well across all datasets. Furthermore, based on the significant performance improvement of the FinBERT model with our Twitter data in general domain compared to our news data in general domain, we also express uncertainty about the model being further pre-trained only on financial news data. We simplify the complex process of determining the optimal approach to the FinBERT model and provide guidelines for selecting additional training datasets and hyperparameters within the fine-tuning process of financial sentiment analysis models.

• Proceedings of the KSRS Conference
• /
• 2006.03a
• /
• pp.127-130
• /
• 2006

SRTM 데이터와 USGS의 NED (National Elevation Datasets) 데이터를 사용하였으며 두 데이터를 차분함으로써 식생수고도(vegetation height map)를 얻었다. 또한 차분값과 shape 파일에 포함된 식수년도의 비교를 통해 상관관계여부를 판단하고자 했다. 회귀분석을 통해 차분데이터와 식수년도 사이의 큰 상관관계가 존재함을 확인할 수 있었으며 결국 수령추정과 수령정보의 맵핑이 가능함을 보였다. 추가적으로 지역별 지형특성, 숲의 균일도 등에 의해 선형성이 영향을 받는지 관찰하였다.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2010.04a
• /
• pp.3-4
• /
• 2010

Map integration usually involves matching the common spatial objects in different datasets. There have been recent studies on object matching using relative location as defined by spatial relationships between the object and its neighbor landmark. Therefore the landmark selection process is an important part of map integration using relative location. In this research, we describe an approach to determine landmarks automatically in different geospatial datasets.

• The Bulletin of The Korean Astronomical Society
• /
• v.45 no.1
• /
• pp.50.1-50.1
• /
• 2020

In this topic, I will introduce a recent work on reconstructing the cosmic density field based on the GAN. I will show the performance of the GAN compared to the traditional Unet architecture. I'd also like to discuss a 3-channels-based 2D datasets for the training to recover the 3D density field. Finally, I will present some performance tests based on the test datasets.

• Atmosphere
• /
• v.27 no.1
• /
• pp.1-16
• /
• 2017

In this study, trends of upper jet stream characteristics (intensity, altitude, latitude, and longitude) over the Asia-North Pacific region during the recent 30 (1979~2008) years were analyzed by using four reanalysis datasets (CFSR, ERA-Int., JRA-55, MERRA). We defined the characteristics of upper jet stream as the averages of mass weighted wind speed, mass-flux weighted altitude, latitude and longitude between 400 and 100 hPa. Due to the vertical averaging of jet stream characteristics, our results reveal a weaker spatial variabilities and trends than previous studies. In general, the four reanalysis datasets show similar jet stream properties (intensity, altitude, latitude and longitude) although the magnitude and trends are slightly different among the reanalysis datasets. The altitude of MERRA is slightly higher than that of others for all seasons. The domain averaged intensity shows a weakening trend except for winter and the altitude of jet stream shows an increasing trend for all seasons. Also, the meridional trend of jet core shows a poleward trend for all seasons but it shows a contrasting trend, poleward trend in the continental area but equatorward trend in the Western Pacific region during summer. The zonal trend of jet core is very weak but a relatively strong westward trend in jet core except for spring and winter. The trends of jet stream characteristics found in this study are thermodynamically consistent with the global warming trends observed in the Asia-Pacific region.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 2008.06a
• /
• pp.310-317
• /
• 2008

Although aerial LiDAR had been launched commercially several years ago, it is still difficult to study data acquisition conditions and effects with various datasets because of its acquisition cost. Thus, this research was performed to study data acquisition conditions and effects with virtually various datasets. For this research, 3D tree models and forest stand models were built to represent graded tree sizes and tree plantation densities. Also, a variable aerial LiDAR acquisition model was developed. Then, through controlling flight height parameter, one of the data acquisition parameters, virtual datasets were collected for various data acquisition densities. From those datasets, forest canopy volumes and maximum tree heights were estimated and the estimated results were compared. As the results, the estimated is getting closer to the expected during the data acquisition density increase. This research would be helpful to perform further studios on relations between forest inventory accuracy and LiDAR cost.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.32 no.2
• /
• pp.87-93
• /
• 2014

Cadastral parcel objects as polygons are fundamental dataset which represent land administration and management of the real world. Thus it is necessary to assure topological seamlessness of cadastral datasets which means no overlaps or gaps between adjacent parcels. However, the problem of overlaps or gaps are frequently found due to non-coinciding edges between adjacent parcels. These erroneous edges are called uncertain edges, and polygons containing at least one uncertain edge are called uncertain polygons. In this paper, we proposed a new algorithm to efficiently search parcels of uncertain polygons between two adjacent cadastral datasets. The algorithm first selects points and polylines around adjacent datasets. Then the Constrained Delaunay Triangulation (CDT) is applied to extract triangles. These triangles are tagged by the number of the original cadastral datasets which intersected with the triangles. If the tagging value is zero, the area of triangles mean gaps, meanwhile, the value is two, the area means overlaps. Merging these triangles with the same tagging values according to adjacency analysis, uncertain edges and uncertain polygons could be found. We have performed experimental application of this automated derivation of partitioned boundary from a real land-cadastral dataset.

• Molecules and Cells
• /
• v.42 no.3
• /
• pp.237-244
• /
• 2019

Understanding the mechanisms of cancer drug resistance is a critical challenge in cancer therapy. For many cancer drugs, various resistance mechanisms have been identified such as target alteration, alternative signaling pathways, epithelial-mesenchymal transition, and epigenetic modulation. Resistance may arise via multiple mechanisms even for a single drug, making it necessary to investigate multiple independent models for comprehensive understanding and therapeutic application. In particular, we hypothesize that different resistance processes result in distinct gene expression changes. Here, we present a web-based database, CDRgator (Cancer Drug Resistance navigator) for comparative analysis of gene expression signatures of cancer drug resistance. Resistance signatures were extracted from two different types of datasets. First, resistance signatures were extracted from transcriptomic profiles of cancer cells or patient samples and their resistance-induced counterparts for >30 cancer drugs. Second, drug resistance group signatures were also extracted from two large-scale drug sensitivity datasets representing ~1,000 cancer cell lines. All the datasets are available for download, and are conveniently accessible based on drug class and cancer type, along with analytic features such as clustering analysis, multidimensional scaling, and pathway analysis. CDRgator allows meta-analysis of independent resistance models for more comprehensive understanding of drug-resistance mechanisms that is difficult to accomplish with individual datasets alone (database URL: http://cdrgator.ewha.ac.kr).