In August 2020, South Korea's data activation legislation was updated, introducing “Guidelines for the Use of Health and Medical data,” thereby extending its reach to the medical sector. These guidelines delinea...
In August 2020, South Korea's data activation legislation was updated, introducing “Guidelines for the Use of Health and Medical data,” thereby extending its reach to the medical sector. These guidelines delineate crucial protocols for individual medical data use, emphasizing anonymization and mandatory data Review Committee oversight. Our research proposes a methodology within the KMbig system for health data utilization, compliant with these stipulations. Through comprehensive analysis, we have developed specific anonymization techniques for various data types, enabling direct validation and swift identifiability risk assessment. The KMbig system's deployment is anticipated to significantly enhance healthcare data efficacy, aligning with the regulatory framework.
This paper reports the effect of a novel artificial neural network architecture for industrial anomaly detection using generative adversarial network (GAN)-based data augmen-tation. We show that GAN-based data augment...
This paper reports the effect of a novel artificial neural network architecture for industrial anomaly detection using generative adversarial network (GAN)-based data augmen-tation. We show that GAN-based data augmentation enhances the performance of end-to-end electric pole anomaly detection. With the convolutional neural network (CNN) hyperparameter search, our method outperforms vanilla CNN and Cutout augmentation by an average of 2.2 % p and 1.6 % p, respectively and has an accuracy of over 88 % for the test dataset.
As most of the Internet of Things (IoT) applications are event-driven, the emergence of the serverless computing paradigm, which is a natural fit for event-driven applications, is promising to host multi-tenant IoT ap...
As most of the Internet of Things (IoT) applications are event-driven, the emergence of the serverless computing paradigm, which is a natural fit for event-driven applications, is promising to host multi-tenant IoT applications. Furthermore, the increasing resource capability of low-cost edge and fog devices provides an opportunity to take advantage of resources available and leads to the edge-fog-cloud computing continuum, which can conduct processing across the entire computing continuum. To identify the necessary adaptations for the serverless computing continuum, we integrate the serverless paradigm in each layer of the computing continuum and investigate performance parameters by running serverless workloads using benchmarks.
In NLG, recent research in Knowledge-Grounded Text Generation aims to refine sentence specificity and naturalness. When generating texts, considering multiple turns within a conversation is considered important becaus...
In NLG, recent research in Knowledge-Grounded Text Generation aims to refine sentence specificity and naturalness. When generating texts, considering multiple turns within a conversation is considered important because it allows models to generate sentences that reflect the context of the conversation. Addressing open-domain conversations, determining the optimal conversation history for training knowledge selection models lacks prior exploration. This study aims to improve KGTG models to effectively handle complex utterances by progressively incorporating more turns. This finding offers a foundational direction for boosting knowledge selection models in text generation.
The current health care adopts smart care driven by data, utilizing multiple-sensor measurements. However, it is not straightforward how one may map the relationship of sensors using traditional machine learning metho...
The current health care adopts smart care driven by data, utilizing multiple-sensor measurements. However, it is not straightforward how one may map the relationship of sensors using traditional machine learning methods alone. This paper introduces a method integrating a Graph Convolutional Network (GCN) with an odor-sensing array to extract the change in odor from respiratory information such as concentrations of Volatile Organic Compounds. This approach measures the differences in odor under different conditions of the subjects (e.g., 1. before and after exercise, 2. during COVID-19 sickness and after recovery) by learning the increasing concentration of gas mixtures from multiple sensors. GCN grasps the relationship between odor sensors' sensitivity and achieves an experimental accuracy rate of 81.6%. Since the graph structure is a scalable permutable domain, other odor-gain labels can potentially form a new feature learning based on this pivot feature learning.
The early detection of defects in mechanical equipment is of paramount importance in the industrial field. Research into data analysis methodologies for extracting valuable data from mechanical equipment has highlight...
The early detection of defects in mechanical equipment is of paramount importance in the industrial field. Research into data analysis methodologies for extracting valuable data from mechanical equipment has highlighted the significant value of these technologies. One such study is the early and accurate detection of anomalies in the acoustic data of machinery. In this paper, we propose an effective technique for the detection and classification of rare event anomalies in the data derived from the rotational noise of automotive motors. MFCC extraction and smoothing techniques were used to select minimal features for optimal performance, and Principal Component Analysis (PCA) was applied to extract salient features. These features are capable of distinguishing between normal and anomalous data. Additionally, an unsupervised learning algorithm was applied to the dataset to differentiate between normal and anomaly data. Experimental results showed that the proposed method can effectively detect sound anomalies with a high accuracy of 99.4% and is also capable of detailed classification of anomalous data.
Similar image retrieval identifies and ranks images from a database based on visual characteristics such as color, texture, and shape to match a given sample image, providing results with the closest resemblance. Henc...
Similar image retrieval identifies and ranks images from a database based on visual characteristics such as color, texture, and shape to match a given sample image, providing results with the closest resemblance. Hence, it's crucial to have a thorough understanding of the content in images. In this paper, we propose a method to utilize structural information based on an image segmentation model. Significantly, the structural information can provide novel insights into aspects of images that conventional dense global descriptors may overlook, as it can clarify the shapes of objects and their backgrounds, offering a complementary perspective. Specifically, our method combines structural information with global descriptors, allowing both the detailed shapes and broader features of images to be captured in a manner that users can control. In the experiments, we evaluate the extent to which this integrated approach enhances the performance of similarity search tasks.
Ultra-wideband (UWB) technology is gaining attention for precise indoor positioning, offering high bandwidth compared to alternatives like Bluetooth Low Energy (BLE) and Wi-Fi. Researchers are exploring its potential ...
Ultra-wideband (UWB) technology is gaining attention for precise indoor positioning, offering high bandwidth compared to alternatives like Bluetooth Low Energy (BLE) and Wi-Fi. Researchers are exploring its potential for practical indoor positioning in real-world scenarios when combined with other techniques. This paper presents an Indoor Positioning System (IPS) based on Time of Flight (ToF) and Kalman filtering using UWB sensors. This work aims to develop an IPS that can be used for practical object and person tracking in real-world scenarios. The process entails increasing the precision of ToF measurements using sensor calibration and computing methods, followed by the use of Kalman filtering to enhance location data by refining real-time measurement. The system underwent evaluation in three distinct scenarios, namely Line-of-Sight (LoS), Obstructed-Line-of-Sight (OLoS), and a real-world simulation where the tag was concealed within the pocket of an individual, achieving average positioning errors of 17.21 cm, 48.27 cm, and 46.17 cm, respectively.
Technical advancements in recent decades have led to generation and collection of much more data at a rapid rate from a wide variety of rich data sources. The popularity of initiates of open data has also encouraged t...
详细信息
ISBN:
(纸本)9781728189246
Technical advancements in recent decades have led to generation and collection of much more data at a rapid rate from a wide variety of rich data sources. The popularity of initiates of open data has also encouraged the sharing of these bigdata so that they have become publicly accessible. Examples of these bigdata include transportation data. Analyzing and mining these big transportation data help users (e.g., commuters, city planners) to take appropriate actions (e.g., making wise decisions), which in turn help building a smarter city. This leads to smartcomputing. Moreover, contents of available big transportation data may vary among cities, which lead to the conceptual modeling to describe at a high level of abstraction the semantics of data analytic and mining software applications on big transportation data. In this paper, we present conceptual modeling and smartcomputing for big transportation data. We illustrate our idea with real-life big transportation data from the Canadian city of Winnipeg and to show its practicality in real-life data.
暂无评论