第116回総合防災セミナー(8/7)

[Date & Time]
8月7日（金）15:00-17:00

[Venue]
宇治キャンパス本館S519D　および　オンライン
※オンラインでご参加の方は8月6日（木）までに、下記リンクよりお申し込みください．
 

登録フォーム：https://forms.gle/37griFTu7fo6YaeF9

 
[Speaker]
Dr. Mohammad Reza Yeganegi
Researcher Scholar,  Cooperation and Transformative Governance Research Group, Advancing Systems Analysis, International Institute for Applied Systems Analysis (IIASA)
Visiting associate professor, DRS, DPRI, Kyoto University

 
[Title]
Inclusive and Evidence-Based Decisions for Resilience building and Disaster Risk Management: Statistical Foundations and AI-powered Models for Numeric and Textual Data Analysis

 
[Abstract]
Effective resilience building and managing disaster risk rely on making the decision based on existing evidence and including different stakeholders in the decision-making process. Part of this evidence is structured quantitative and qualitative data. The other part of the information is experts’ opinions and judgments. Public communications, narratives, and discourse are another source of data that contains local knowledge, communities’ concerns, and priorities. Designing such a decision framework requires the methodology for integrating various quantitative and qualitative factors. AI-driven models have been used for analyzing massive amounts of data and providing input to decision models. Additionally, the surge of advancements in deep learning methods and language models offers a unique opportunity to analyze a massive amount of unstructured data. Despite these advancements, the challenges remain in integrating different data sources into a risk-based decision model and extracting the necessary information from textual data to include in the decision-making process. This talk will focus on the statistical foundations necessary for building such decision frameworks and explore the existing AI-based models, along with the challenges of building such decision frameworks.

 
[Biography]
Mohammad Reza Yeganegi is research scholar at IIASA’s Advancing Systems Analysis (ASA) Program. He has a PhD in Statistics, with a focus on the theory of dynamic systems and time series analysis. Prior to joining IIASA, he was assistant professor in financial statistics and led the courses in advanced financial statistics and financial econometrics. During this time, he worked on nonlinear time series and dynamic systems.
Since joining IIASA, Yeganegi has been leading the development of natural language processing methodology and models for narrative analysis, social media mining tools. This methodology is used for understanding public concerns and narratives. Additionally, he is developing impact-based decision models. He has worked on developing human-centered decision analysis framework for disaster risk management. His research is focusing on practical narrative analysis and frameworks that integrates quantitative measures with insights from public narratives and unstructured data into an impact-based decision.