AxeLatoon - Let's Build an Accelerator at School !

High Energy Accelerator Research Organization (KEK) launched an education project for the fabrication of an accelerator named "AxeLatoon" in 2020 together with the National Institute of Technology (KOSEN). This project aims to improve engineering skills of students and foster the next generation of accelerator researchers by providing hands-on training in the field of accelerator science. In the first year, we collaborated with the NIT (KOSEN), Ibaraki College to build an accelerator. Students took the initiative in this extracurricular activity and challenged building an accelerator. From 2021, we expanded this project to other prefectures and four schools are now participating. The design and fabrication of a small cyclotron accelerator is currently underway. Despite the restrictions on activities and the limited mobility of people due to the novel coronavirus pandemic, the project continues to educate students about basic technologies and accelerators. We are holding seminars a few times a month utilizing online communication tools. In this report, we would like to share the status of AxeLatoon's activities based on the actual production of students at KOSEN and deepen the discussion on accelerator outreach programs. © Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0)

Weighted Graph-Based Signal Temporal Logic Inference Using Neural Networks

Extracting spatial-temporal knowledge from data is useful in many applications. It is important that the obtained knowledge is human-interpretable and amenable to formal analysis. In this paper, we propose a method that trains neural networks to learn spatial-temporal properties in the form of weighted graph-based signal temporal logic (w-GSTL) formulas. For learning w-GSTL formulas, we introduce a flexible w-GSTL formula structure in which the user’s preference can be applied in the inferred w-GSTL formulas. In the proposed framework, each neuron of the neural networks corresponds to a subformula in a flexible w-GSTL formula structure. We initially train a neural network to learn the w-GSTL operators, and then train a second neural network to learn the parameters in a flexible w-GSTL formula structure. We use a COVID-19 dataset and a rain prediction dataset to evaluate the performance of the proposed framework and algorithms. We compare the performance of the proposed framework with three baseline classification methods including K-nearest neighbors, decision trees, support vector machine, and artificial neural networks. The classification accuracy obtained by the proposed framework is comparable with the baseline classification methods. IEEE