| 山内 修平 | D, 中間発表 | 情報基盤システム学 | 藤川 和利, 笠原 正治, 柴田 直樹 |
|
title: Research on Preventing Double-Counting in Carbon Offsetting through the Use of Blockchain and NFTs
abstract: Carbon offsetting is widely adopted as part of climate change mitigation efforts; however, the issue of double-counting has been raised. This refers to the same carbon offset being accounted for more than once, resulting in an overestimation of actual reduction efforts. The aim of this study is to develop a method to address the double-counting issue by constructing a system that leverages blockchain technology and Non-Fungible Tokens (NFTs). The system tracks carbon offsets generated at each stage of the manufacturing process and verifies the carbon emission data registered on NFTs. It also has the capability to integrate multiple NFTs to evaluate the overall carbon offset of a product. Through this system, the reliability of publicized NFTs can be confirmed, and carbon offsets can be tracked and verified, preventing double-counting. language of the presentation: Japanese 発表題目: ブロックチェーンを活用したNFTによるカーボンオフセットのダブルカウンティング防止システムの研究 発表概要: カーボンオフセットは気候変動対策の一環として広く採用されているが、ダブルカウンティングの問題が指摘されている。これは、同じカーボンオフセットが二重に計算され、その結果、実際の削減量が過大評価される問題である。本研究の目的は、ブロックチェーン技術と非代替性トークン（NFT）を活用したシステムを構築し、このダブルカウンティング問題に対処する手法を開発することである。システム は、製造過程の各段階で発生するカーボンオフセットを追跡し、NFTに登録されたカーボン排出量データを検証する。そして複数のNFTを統合することで製品全体のカーボンオフセットを評価する機能を持つ。このシステムにより、公開されたNFTの信頼性が確認でき、ダブルカウンティングを防止しながら、カーボンオフセットの追跡と検証が可能となる。 | |||
| 岸下 昂生 | M, 2回目発表 | 数理情報学 | 池田 和司, 笠原 正治, 久保 孝富, 日永田 智絵 |
|
title: Development of Defecation Prediction System Using Sequential Bayesian Estimation with Bowel Sounds
abstract: As the body ages, bowel motor and sensory functions decline. This increases the risk of defecation problems. The burden caused by defecation disorders falls not only on the elderly themselves, but also on care workers and nurses. Therefore, this study aims to develop a system to detect defecation using bowel sounds. In a previous study, they calculated the signal power of sounds for about 30 seconds before and after defecation and showed the possibility of predicting defecation based on differences in their distribution. The final goal of this research is to develop an AI system that sequentially measures and extracts features of bowel sounds using a wearable device and sequentially updates the time of defecation. The master's research aims to develop an AI system that actually predicts defecation sequentially by recording bowel sounds with a non-wearable stethoscope. At present, we have (1) developed a recording environment and (2) extracted features from data from previous studies, and demonstrated the possibility of predicting bowel movements using machine learning algorithms other than those used in previous studies. As a future prospect, we aim to develop a system to predict the time of defecation from bowel sounds by collecting data sets of bowel sounds and time of defecation for several people. language of the presentation: Japanese | |||
| KAUSMALLY MOHAMMAD SHAHOOR HUSAIN | M, 2回目発表 | 数理情報学 | 池田 和司, 松本 健一, 久保 孝富, 日永田 智絵 |
|
title: Functional categories of code links between neural efficiency and expertise level
abstract: In recent decades, programming has become an important skill, drawing considerable interest from researchers examining how the human brain handles it. This study investigates whether programmers' brains work more efficiently based on their expertise. To explore this, we use functional Magnetic Resonance Imaging (fMRI) data collected in a previous study and analyze the size of activated brain areas. Our findings show that each programmer has a unique way of processing programming tasks, and this remains consistent regardless of their expertise level. This suggests that the brain's efficiency is influenced by factors beyond expertise alone. Additionally, it hints that the brain might treat different types of code similarly. language of the presentation: English | |||
| 相島 祐太 | M, 2回目発表 | 数理情報学 | 池田 和司, 松原 崇充, 久保 孝富, 日永田 智絵 |
|
title: Transport Analysis for Regression
abstract: Recent theoretical studies have shown that the denoising autoencoder(DAE) is a mapping that transports the data distribution in a direction that decreases its entropy. In this study, we extend this theory and perform a transport theoretic analysis of the regression problem. First, a transport map is derived from the optimal solution of the regression. We confirm that this transport map works in the same way as the DAE transport map. Furthermore, we show the existence of another transport map and discuss the relationship between the transport map and the transport map derived from the optimal solution. language of the presentation: *** English or Japanese (choose one) *** | |||