ゼミナール発表

日時: 6月1日(水)3限 (13:30-15:00)


会場: L1

司会:川波 弘道
Afendi Farit Mochamad D 金谷重彦 小笠原直毅 Md.Altaf-Ul-Amin 中村建介 高橋弘喜
発表題目:Statistical Model of Plants Function in Jamu Efficacy
発表概要:Jamu is Indonesian herbal medicine made from a mixture of several plants. The plants are chosen mainly by experience; pass down from generation to generation. Some plants perform as main ingredients in jamu, which cure the disease directly; whereas some others act as supporting ingredients, which help curing the disease indirectly. However, which plants act as main ingredients and which for supporting ingredients need to be further determined. This research is aimed to determine the function of plants in jamu efficacy by means of statistical models. We proposed Biplot to obtain relationship between plants and jamu efficacy; whereas PLS-DA model is used to predict jamu efficacy using information of its ingredients. Furthermore, Multi-way model will be used in modeling plants usage in jamu along with their pharmacological activities to predict the efficacy of jamu.
 
Aisuwarya Ratna M1回目 知能コミュニケーション
Multilingual Acoustic Modelling for Highly Accented Indonesian ASR
Indonesia in particular, as the fourth most populous nation in the world still lacks speech technology and research, and also suffers from a shortage of Indonesian language data. Cross-language approach technique between English and Indonesian (Sakti, 2005) have been proposed, and show significant results for non-accented Indonesian Speech Recognition. Thus, extend research is needed in order to improve recognition performance for accented Indonesian speakers, due to the reason that most of the peoples speaks local languages as their first language, which affects their proficiency in speaking Indonesian. We will try to investigate the possibility to build acoustic model based on multilingual approach. .
 
藤田 裕介 M1回目 インタラクティブメディア設計学