Kyushu Institute of Information Sciences

There have been reports suggesting the possibility of the West Pacific Anomaly, an area of weak magnetic fields, existing around Southeast Asia during the 15–19 th century. To verify whether such an anomaly existed in other ages, we conducted a study

Abstract While waterlogging stress slows photosynthetic rate ( A sat ), the underlying processes remain poorly understood. Here, we aimed to characterize the limitations to photosynthesis imposed by stomatal conductance ( g s ), mesophyll conductance ( g m ), and biochemical processes under waterlogging and subsequent reoxygenation. Two soybean cultivars ( Glycine max L. cv. Fukuyutaka and Iyodaizu) were subjected to 6 days of waterlogging, after which excess water was drained. The responses of A sat , g s , g m , and the maximum carboxylation rate ( V cmax ) were investigated. In both cultivars, A declined significantly within 4 days of waterlogging and did not recover completely by two weeks of reoxygenation. During waterlogging, CO 2 concentration at carboxylation site decreased in parallel with g s and g m , indicating that photosynthesis was mainly limited by diffusional factors (combination of g s and g m ). After drainage, diffusional limitation persisted during early reoxygenation, whereas biochemical limitation due to reduced V cmax became dominant after 7 days of reoxygenation. Therefore, maintaining high diffusional conductances and V cmax during waterlogging and reoxygenation, respectively, is important for enhancing photosynthetic tolerance to waterlogging stress. Overall, our results demonstrate that A sat under waterlogging and reoxygenation is dynamically constrained by multiple factors, emphasizing the need for comprehensive assessment of gas diffusion and carbon assimilation processes.

We investigate the high-latitude Joule heating in a recently developed high-resolution configuration of the WACCM-X whole-atmosphere model coupled to the GAMERA magnetosphere model. Model resolution and geomagnetic forcing effects during the severe ge

Previously we proposed a CAPTCHA using images processed with a weave filter. This was effective in preventing object recognition AI from correctly recognizing objects. However, it was unclear why it could prevent AI from recognizing objects. In this paper, we discuss through experiments why the object detection AI cannot correctly recognize weaved images. We input images with weave processing and images with similar masks applied to YOLOv5, which is an object detection AI, and observe the results. We hypothesize the reason why the recognition results differ between the different processed images, by considering the processing features. Based on the results, we will conduct further experiments to confirm that YOLOv5 considers outlines and some parts of the image as important features for recognition. Finally, we will provide future work on the generation of CAPTCHA images.

We study a 4-dimensional Lorenz-Stenflo system for generating random bit sequences shared between two entities. In this study, we present an integerized Lorenz-Stenflo system that gives the same output even when implemented on different CPU architectures and has an advantage of requiring less calculation than floating-point arithmetic. Experimentally we show that our proposal can synchronize faster than floating-point implementations.

In neuroscience-inspired computing, reservoir computing (RC) has emerged as a powerful model that leverages recurrent dynamics to mimic neural systems. Echo State Networks (ESNs), a prominent RC framework, are effective for modeling temporal data. However, physical RC using pulse-coupled phase oscillator (PCPO) networks, while promising for hardware systems, often struggle with flexibility, limiting their performance in time series forecasting. To address these challenges, this paper introduces a novel hybrid RC framework that combines the both states from ESNs and PCPO using two oscillators towards optimization physical RC. This hybrid-state using the oscillating events and enhance the representational capacity of the reservoir, thereby enabling the model to learn pertinent features for more accurate predictions. We evaluate our framework using the synthetic NARMA-10 dataset and the real-world Numenta Anomaly Benchmark (NAB). Experimental results show that the proposed method achieves lower mean squared error (MSE) values 7.19×10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−3</sup> on NARMA-10 and 33.60×10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−3</sup> on NAB, which outperforming the baseline ESN model.

This paper introduces a novel nonlinear model predictive control (NMPC) framework that incorporates a lifting technique to enhance control performance for nonlinear systems. While the lifting technique has been widely employed in linear systems to capture intersample behaviour, their application to nonlinear systems remains unexplored. We address this gap by formulating an NMPC scheme that combines fast-sample fast-hold (FSFH) approximations and numerical methods to approximate system dynamics and cost functions. The proposed approach is validated through two case studies: the Van der Pol oscillator and the inverted pendulum on a cart. Simulation results demonstrate that the lifted NMPC outperforms conventional NMPC in terms of reduced settling time and improved control accuracy. These findings underscore the potential of the lifting-based NMPC for efficient control of nonlinear systems, offering a practical solution for real-time applications.

One of the resting state networks is the default mode network (DMN); the DMN is primarily concerned with self-perception, orientation, and memory. Although EEG-DMN has been suggested to be active in the ultra-low frequency band below 1 Hz based on fMRI studies, the characteristics of the network have not yet been elucidated. In this study, we measured EEG-DMN during the eye-closed resting state(EC), eye-open resting state(EO), and mental arithmetic with eye-close(CA), and analyzed network activity using phase locking value(PLV), an index of phase-locking, which is one of the functional connectivity. The results showed that functional connectivity between regions (frontal, central, posterior, occipital) including DMN-related regions in SubDelta (0.01-0.1Hz) was significantly higher during eye-closed resting state than during mental arithmetic with eye-close (EC PLV - CA PLV in Frontal-Posterior = 0.062). In contrast, in the high frequency band(Theta), PLV during mental arithmetic with eye-close was significantly higher than during eye-closed resting between Occipital-Right temporal (EC PLV – CA PLV in Posterior-Right temporal = -0.051). This finding is similar to previous studies and suggests that the analysis of functional connectivity in the very low-frequency band below 0.1 Hz can reveal EEG-DMN activity. This result indicates that DMN activity can be captured not only by fMRI and intracranial EEG (iEEG) but also by cortical EEG. This will contribute to the elucidation of the pathology of diseases with reported abnormal DMN activity, such as depression and dementia, as well as to the further elucidation of DMN function.

High school baseball players across the country are practicing hard every day with the goal of playing at Koshien. In recent years, data analysis has become the key to determining the outcome of games. So we developed this system to contribute to Koshien participation and to increasing the number of baseball players. This time, we conducted a simple regression analysis to examine the correlation between batting average and strikeout rate. A simple regression analysis revealed that the contribution rate was 8%, indicating no correlation between the two.