【目的】災害医療対応シミュレーション・システムを用いて，首都直下地震発災後の人的被害や医療資源を想定し，効果的な医療活動を行うために有効な施策を検討した。【対象と方法】東京都ガイドラインの傷病者フローに従った。発生傷病者数，病床数，保健医療活動チーム数，域内搬送用救急車数，広域医療搬送用回転翼機数の数値を変化させた際の発災後死亡者数への影響を検討した。【結果および考察】発生傷病者数が想定の21,511人では発災後死亡者数は6,641人（30.9%）であるが，想定の0.8倍では3,934人（25.3%），0.6倍では1,919人（14.5%）と減少し，0.4倍であれば，137名（1.6%）まで減少させることが示された。病床数は，それのみでは発災後死亡減少への寄与は僅かであることがわかった。保健医療活動チームは，派遣チーム数を増やすことができれば，死亡者が減少することがわかった。救急車や回転翼機数の発災後死亡者数減少への寄与は少ないことが示された。今回の検討は，あくまでもシミュレーション上の結果であり，現実を反映したものではない。【まとめ】首都直下地震を想定し，発生傷病者数，病床数，保健医療活動チーム数，救急車数，回転翼機数を検討した結果，傷病者数の減少，保健医療活動チーム数の増加により発災後死亡は減少する可能性が示唆された。今後は発生傷病者数を減らすための防災・減災努力を行うとともに，災害医療対応のベストミックスを検討する必要がある。

The usefulness and value of Multi-step Machine Learning (ML), where a task is organized into connected sub-tasks with known intermediate inference goals, as opposed to a single large model learned end-to-end without intermediate sub-tasks, is presented. Pre-optimized ML models are connected and better performance is obtained by re-optimizing the connected one. The selection of an ML model from several small ML model candidates for each sub-task has been performed by using the idea based on Neural Architecture Search (NAS). In this paper, Differentiable Architecture Search (DARTS) and Single Path One-Shot NAS (SPOS-NAS) are tested, where the construction of loss functions is improved to keep all ML models smoothly learning. Using DARTS and SPOS-NAS as an optimization and selection as well as the connections for multi-step machine learning systems, we find that (1) such a system can quickly and successfully select highly performant model combinations, and (2) the selected models are consistent with baseline algorithms, such as grid search, and their outputs are well controlled.

We study the possibilities of a search for the light and weakly interacting gauge boson in the gauged $L_\mu - L_\tau$ model. Introducing the kinetic mixing at the tree level, the allowed parameter regions for the gauge coupling and kinetic mixing parameter are presented. Then, we analyze one photon plus missing event within the allowed region and show that a search for the light gauge boson will be possible at the Belle-II experiment. We also analyze the neutrino trident production process in neutrino beam experiments.

We discuss the Dirac CP violating phase $\delta_{CP}$ in the Froggatt–Nielsen model for a neutrino mass matrix $M_\nu$ imposing a condition $\det[M_\nu]=0$. This additional condition restricts the CP violating phase $\delta_{CP}$ drastically. We find that the phase $\delta_{CP}$ is predicted in the region of $\pm(0.4–2.9)$ radian, which is consistent with the recent T2K and NO$\nu$A data. There is a remarkable correlation between $\delta_{CP}$ and $\sin^2\theta_{23}$. The phase $\delta_{CP}$ converges to $\sim\pi/2$ if $\sin^2\theta_{23}$ is larger than 0.5. Thus, the accurate measurement of $\sin^2\theta_{23}$ is important for a test of our model. The effective mass $m_{ee}$ for the neutrinoless double beta decay is predicted in the rage $3.3–4.0$ meV.

The model of neutrino mass matrix with minimal texture is now tightly constrained by experiment so that it can yield a prediction for the phase of CP violation. This phase is predicted to lie in the range $\delta_{CP}=0.77\pi–1.24\pi$. If neutrino oscillation experiment would find the CP violation phase outside this range, this means that the minimal-texture neutrino mass matrix, the element of which is all real, fails and the neutrino mass matrix must be complex, i.e., the phase must be present that is responsible for leptogenesis.