Kyutacar : Kyushu Institute of Technology Academic Repository
Not a member yet
9746 research outputs found
Sort by
Analysis of Obese Patients Using Machine Learning to Categorize Hidden Risk Factors in Explorative Assessment
Get PDFObesity has a multitude of detrimental impacts on both mental and physical health, along with an increased risk of acquiring chronic illnesses including coronary heart disease, diabetes, and stroke, as well as anxiety and depression conditions. Using forecasting analytics, researchers may develop therapies and prevention methods that are tailored to a particular behavior and medical knowledge. Data analytics examines the greater risk of contracting diseases associated with obesity in order to target a specific group. In order to undertake efficient treatments, it is primarily important to gather and evaluate data on issues associated with obesity. Gathering the abnormal diet, lifestyle, and previous medical history tends to indicate whether the individual is affected by an obese hidden risk factor. Machine learning can assist in early diagnosis and anticipate potential health problems associated with obesity. The proposed approach involves various machine learning techniques to analyze comprehensive data on diet, lifestyle.journal articl
Hybrid IoT Programming with Data Flow Diagrams and mruby
Get PDFIoT is increasingly being used in different areas of society. IoT is used in society by collecting values acquired by sensor devices through networks and analysing the data. IoT software is needed to use the IoT, but efficient software development methods are desired because software for an increasingly complex society tends to become more complex. This research aims to improve the efficiency of IoT software and develop an intuitive method for developing IoT software.This paper proposes a development methodology that combines a data-flow-based programming approach with a procedural programming language. When the focus is on data, data flow programming is intuitive and efficient. On the other hand, when the focus is on data processing, it is better to use a procedural programming language because data is processed sequentially[1].In the hybrid IoT programming environment proposed in this paper, the code of a procedural programming language is generated from the description by data flow programming, and the program is executed by a micro controller. The products of this research are available as open source software[2], [3].journal articl
NIR-Sensitive Squaraine Dye—Peptide Conjugate for Trypsin Fluorogenic Detection
Get PDFTrypsin enzyme has gained recognition as a potential biomarker in several tumors, such as colorectal, gastric, and pancreatic cancer, highlighting its importance in disease diagnosis. In response to the demand for rapid, cost-effective, and real-time detection methods, we present an innovative strategy utilizing the design and synthesis of NIR-sensitive dye–peptide conjugate (SQ-3 PC) for the sensitive and selective monitoring of trypsin activity by fluorescence ON/OFF sensing. The current research deals with the design and synthesis of three unsymmetrical squaraine dyes SQ-1, SQ-2, and SQ-3 along with a dye–peptide conjugate SQ-3-PC as a trypsin-specific probe followed by their photophysical characterizations. The absorption spectral investigation conducted on both the dye alone and its corresponding dye–peptide conjugates in water, utilizing SQ-3 and SQ-3 PC respectively, reveals enhanced dye aggregation and pronounced fluorescence quenching compared to observations in DMSO solution. The absorption spectral investigation conducted on dye only and corresponding dye–peptide conjugates in water utilizing SQ-3 and SQ-3 PC, respectively, reveals not only the enhanced dye aggregation but also pronounced fluorescence quenching compared to that observed in the DMSO solution. The trypsin-specific probe SQ-3 PC demonstrated a fluorescence quenching efficiency of 61.8% in water attributed to the combined effect of aggregation-induced quenching (AIQ) and fluorescence resonance energy transfer (FRET). FRET was found to be dominant over AIQ. The trypsin-mediated hydrolysis of SQ-3 PC led to a rapid and efficient recovery of quenched fluorescence (5-fold increase in 30 min). Concentration-dependent changes in the fluorescence at the emission maximum of the dyes reveal that SQ-3 PC works as a trypsin enzyme-specific fluorescence biosensor with linearity up to 30 nM along with the limit of detection and limit of quantification of 1.07 nM and 3.25 nM, respectively.journal articl
Peak detection for current balancing of parallel-connected SiC power devices using PCB sensors
Get PDFThe parallelling of power semiconductor devices is essential for desired current ratings and the system is prone to the current unbalancing due to parameter variations. SiC devices with significant variable threshold voltage due to manufacturing yield and temperature distribution have a consequential possibility of dynamic current unbalancing. This paper presents the peak detection-based current balancing of parallel-connected SiC devices to minimize the turn-on and turn-off current unbalancing. PCB current sensors are used for the measurement and feedback of the signal for this peak detection-based current balancing mechanism.conference pape
Formulation of the JSiSE Code of Ethics
Get PDFJSiSE Code of Ethics was newly established at the General Meeting of Members held in June of this year. In this review, we first introduce the process of developing the Code of Ethics for consideration by the Code of Ethics Committee. Next, we will explain what it means for an academic society to establish a Code of Ethics, covering several topics. Finally, the basic principles of the Code of Ethics and the intent behind each article will be described.journal articl
Q-Learning with Prior Knowledge
Get PDF強化学習は報酬が高い行動を学習する方法であり，その代表的アルゴリズムの一つがQ-学習である．Q-学習においてregretの解析が進んできたが，多くの分析は一から学習するということを前提としている．もし，学習済みの知識を活かすことができれば，転移学習などの研究で示されているように改善が可能である．そこで，本研究では事前に学習対象に対する知識を得られるという仮定の下でのQ-学習アルゴリズムBiased Exploration Q-learning (BEQ) を提案する．そしてBEQのregretの解析を行い，regretの上界がO(ˆmax√2′′ + ′′Δ) であり，知識を使っているという違いがあるものの，既存手法より小さくなることを示す．また，実験を通じて，事前知識を導入する既存手法Potential Based Reward Shapingに対して，BEQは優れていることを示す．Reinforcement learning (RL) is a method for learning actions which are highly rewarding and one of the representative methods of RL is Q-learning. Regret of Q-learning have been analyzed, however most of them assume that learning takes place from scratch. If the agent has acquired prior knowledge about the learning domain, its learning will be more efficient as suggested in transfer learning research. Therefore, we propose a Q-learning method, Biased Exploration Q-learning (BEQ), which assumes that the agent can acquire domain knowledge in advance. We analyze regret of BEQ and show that its upper bound is O(ˆmax√2′′+′′Δ). BEQ and existing methods differ in terms of whether the agent can obtain the domain knowledge or not, but this bound is smaller than those of existing methods. Also experiments show that BEQ outperforms Potential Based Reward Shaping (PBRS), which is a common method for introducing domain knowledge.conference pape
Diffusion Action Segmentationにおけるフレーム特徴量へのマスキングに関する研究
Get PDF動画像を理解するための技術としてAction Segmentation（行動分割）がある. 行動分割の研究の一つとして拡散モデルを用いたDiffusion Action Segmentation[1] がある．文献[1] では，人間の行動に基づいたマスキングが設計されており，一つの入力動画像に対して一つのマスキングを選択し，学習する．本研究では，マスキングが選択される比率について調査し，マスキング選択比率最適化のためのアルゴリズムを提案することで，設計者の選択比率の設定にかかる負担を軽減可能にすることを示す．conference pape
Astronomical 6U CubeSat mission VERTECS: Scientific objective and project status
Get PDFWe describe scientific objective and project status of an astronomical 6U CubeSat mission VERTECS (Visible Extragalactic background RadiaTion Exploration by CubeSat). The scientific goal of VERTECS is to reveal the star-formation history along the evolution of the universe by measuring the extragalactic background light (EBL) in the visible wavelength. Earlier observations have shown that the near-infrared EBL is several times brighter than integrated light of individual galaxies. As candidates for the excess light, first-generation stars in the early universe or low-redshift intra-halo light have been proposed. Since these objects are expected to show different emission spectra in visible wavelengths, multi-color visible observations are crucial to reveal the origin of the excess light. Since detection sensitivity of the EBL depends on the product of the telescope aperture and the field of view, it is possible to observe it with a small but wide-field telescope system that can be mounted on the limited volume of CubeSat. In VERTECS mission, we develop a 6U CubeSat equipped with a 3U-sized telescope optimized for observation of the visible EBL. The bus system composed of onboard computer, electric power system, communication subsystem, and structure is based on heritage of series of CubeSats developed at Kyushu Institute of Technology in combination with high-precision attitude control subsystem and deployable solar array paddle required for the mission. The VERTECS mission was selected for JAXA-Small Satellite Rush Program (JAXA-SMASH Program), a new program that encourages universities, private companies and JAXA to collaborate to realize small satellite missions utilizing commercial small launch opportunities, and to diversify transportation services in Japan. We started the satellite development in December 2022 and plan to launch the satellite in FY2025.journal articl
Micromagnetic analysis of magnetic vortex dynamics for reservoir computing
Get PDFReservoir computing (RC) has generated significant interest for its ability to reduce computational costs compared to traditional neural networks. The performance of the RC element is quantified by its memory capacity (MC) and prediction capability. In this study, we utilize micromagnetic simulations to investigate a magnetic vortex based on a permalloy ferromagnetic layer and its dynamics in RC. The nonlinear dynamics of the vortex core (VC), driven by continuous oscillating magnetic fields and binary digit data as spin-polarized current pulses, are analyzed. The highest MC observed is 4.1, corresponding to the nonlinear VC dynamics. Additionally, the prediction capability is evaluated using the Nonlinear Auto-Regressive Moving Average 2 task, demonstrating a normalized mean squared error of 0.0241 highlighting the time-series data prediction performance of the vortex as a reservoir.journal articl
Standardized Language Tests: An Opportunity for Teacher Development?
Get PDFScholars have long espoused the value that language learning experience can have for language teachers. Additionally, in modern language education, standardized tests are seen as critically important. However, no research has been performed in the intersection of these areas – how does taking a standardized test in an additional language influence language teachers? An interview-based study was designed to explore this gap in existing literature, aiming to illuminate how the test-taking experience influences teacher methodologies and student empathy. While the benefits reported by participants in this interview study are highly varied and individualized, all participants claimed that the experience helped them grow as educators and deepened their understanding of the student experience, suggesting that going through the process of a standardized language test may be beneficial for language teachers.journal articl